From fukuta.saori at jp.fujitsu.com Thu Mar 1 06:46:34 2007 From: fukuta.saori at jp.fujitsu.com (Saori Fukuta) Date: Thu, 01 Mar 2007 15:46:34 +0900 Subject: [et-mgmt-tools] [PATCH] waiting VNC port number take2 In-Reply-To: <20070223120004.GA14993@redhat.com> References: <20070223105543.DAE7.FUKUTA.SAORI@jp.fujitsu.com> <20070223120004.GA14993@redhat.com> Message-ID: <20070301154622.B2DC.FUKUTA.SAORI@jp.fujitsu.com> Hi, On Fri, 23 Feb 2007 12:00:05 +0000 "Daniel P. Berrange" wrote: > > > - Once the above fix in libvirt is done, we can make virt-install do > > > a loop like the one you have below, but instead of looking in xenstore > > > it will look at the port in the XML - waiting for it to change from '-1' > > > to an actual port number > > > > You are right, I will consider again as you pointed out. > > Excellant. I made a patch for current libvirt(revision 1.420) that fixed following: revision 1.417 date: 2007/02/27 15:22:13; author: berrange; state: Exp; lines:+7 -0 Don't hardcode port=5900+domid for new xend Signed-off-by: Saori Fukuta Thanks, Saori Fukuta. Index: virt-install =================================================================== diff -r 1776ef836bf6 virt-install --- a/virt-install Wed Feb 28 07:47:31 2007 -0500 +++ b/virt-install Thu Mar 01 23:26:17 2007 +0900 @@ -313,11 +313,19 @@ def get_xml_string(dom, path): return None def vnc_console(dom): - import time; time.sleep(2) # FIXME: ugh. - vncport = get_xml_string(dom, - "/domain/devices/graphics[@type='vnc']/@port") - if vncport == None: - vncport = 5900 + dom.ID() + import time; + num = 0 + while num < ( 40 / 0.25 ): # 40 seconds, .25 second sleeps + vncport = get_xml_string(dom, + "/domain/devices/graphics[@type='vnc']/@port") + if vncport == '-1': + num += 1 + time.sleep(0.25) + else: + break + if vncport == '-1' or vncport is None: + print >> sys.stderr, "Unable to connect to graphical console; vncport number is not found." + return None vncport = int(vncport) vnchost = "localhost" if not os.path.exists("/usr/bin/vncviewer"): From dhawal at netmagicsolutions.com Thu Mar 1 07:22:27 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Thu, 01 Mar 2007 12:52:27 +0530 Subject: [et-mgmt-tools] Released -- Cobbler 0.4.3, Koan 0.2.6 In-Reply-To: <45E60641.5060209@redhat.com> References: <45E60641.5060209@redhat.com> Message-ID: <45E67F33.1060409@netmagicsolutions.com> Michael DeHaan wrote: > Primarily bugfixes. > Cobbler 0.4.3 > - Added netboot_enabled option for systems to control install loops in > programmatic context. (you can ignore this, it's not surfaced in the CLI) > - Disabling anchors in YAML serialization (which make files harder to edit) > - Fix bug in ksmeta argument processing, takes whitespace again, not commas > - Fix bug in old-style deserialization with str and int concatenation > > Koan 0.2.6 > - Store images in a SELinux friendly path > - the output of --list-profiles is now sorted > > Out on et.redhat.com now and waiting on the Fedora mirrors... Just upgraded to 0.4.3, no issues out of the box to report.. will let you know post some testing.. - dhawal From fj7716hz at aa.jp.fujitsu.com Thu Mar 1 08:11:28 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Thu, 01 Mar 2007 17:11:28 +0900 Subject: [et-mgmt-tools] [PATCH]Check the starting domain's uuid Message-ID: <45E68AB0.9070908@aa.jp.fujitsu.com> Hi The virt-install command has a problem that the information of the existent domain is overwritten by virt-install command specifing the domain's uuid. Futhermore, the existent domain's set-up file is overwritten on the xen-3.0.4. I make a patch checking the starting domain's uuid on Guest.py. Signed-off-by: Tatsuro Enokura Thanks, Tatsuro Enokura ======================================================================= diff -r 1776ef836bf6 virtinst/Guest.py --- a/virtinst/Guest.py Wed Feb 28 07:47:31 2007 -0500 +++ b/virtinst/Guest.py Fri Mar 02 00:07:29 2007 +0900 @@ -554,7 +554,19 @@ class Guest(object): def _set_defaults(self): if self.uuid is None: - self.uuid = util.uuidToString(util.randomUUID()) + while 1: + self.uuid = util.uuidToString(util.randomUUID()) + try: + if self.conn.lookupByUUIDString(self.uuid) is not None: + continue + except libvirt.libvirtError: + break + else: + try: + if self.conn.lookupByUUIDString(self.uuid) is not None: + raise RuntimeError, "UUID must be specified for all guests!" + except libvirt.libvirtError: + pass if self.vcpus is None: self.vcpus = 1 if self.name is None or self.memory is None: ======================================================================= From james at cloud9.co.uk Thu Mar 1 09:32:18 2007 From: james at cloud9.co.uk (James Fidell) Date: Thu, 01 Mar 2007 09:32:18 +0000 Subject: [et-mgmt-tools] Xen domU + koan In-Reply-To: <45E4BDA7.3030908@redhat.com> References: <45E43022.1050208@cloud9.co.uk> <45E4504E.4000806@redhat.com> <45E45D61.8080309@cloud9.co.uk> <45E4611C.6060406@redhat.com> <45E4B68B.9010305@cloud9.co.uk> <45E4BDA7.3030908@redhat.com> Message-ID: <45E69DA2.5000908@cloud9.co.uk> Michael DeHaan wrote: > koan needs to be ported over to use the same virtinst-devel python > library that virtguest-install uses, and when that happens, it will be > capable of doing fullvirt installs (and anything else virtinst can > install, which can include virtualization technology other than Xen -- I > believe QEmu is the one virt-manager is supporting now). > > I've wanted to do this earlier -- though I doubt that will happen prior > to mid March. Should someone be feeling adventurous, patches would > definitely be accepted. If not, I'll get around to it when I can -- > it's next up on the cobbler/koan feature list and should require > modifications to koan only, not cobbler. I'm imagining koan will take > a "--fullvirt" flag in addition to it's "--virt" one it takes now. I'd love to do this. Unfortunately I don't yet speak python, so I'm at a bit of a disadvantage. Something else to add to my list of things to do in "my copious free time" :) James From berrange at redhat.com Thu Mar 1 13:51:49 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 1 Mar 2007 13:51:49 +0000 Subject: [et-mgmt-tools] Re: [PATCH] waiting VNC port number take2 In-Reply-To: <20070301154622.B2DC.FUKUTA.SAORI@jp.fujitsu.com> References: <20070223105543.DAE7.FUKUTA.SAORI@jp.fujitsu.com> <20070223120004.GA14993@redhat.com> <20070301154622.B2DC.FUKUTA.SAORI@jp.fujitsu.com> Message-ID: <20070301135149.GA6079@redhat.com> On Thu, Mar 01, 2007 at 03:46:34PM +0900, Saori Fukuta wrote: > Hi, > > On Fri, 23 Feb 2007 12:00:05 +0000 "Daniel P. Berrange" wrote: > > > > > - Once the above fix in libvirt is done, we can make virt-install do > > > > a loop like the one you have below, but instead of looking in xenstore > > > > it will look at the port in the XML - waiting for it to change from '-1' > > > > to an actual port number > > > > > > You are right, I will consider again as you pointed out. > > > > Excellant. > > I made a patch for current libvirt(revision 1.420) that fixed following: > revision 1.417 > date: 2007/02/27 15:22:13; author: berrange; state: Exp; lines:+7 -0 > Don't hardcode port=5900+domid for new xend > > Signed-off-by: Saori Fukuta Great - this patch looks good :-) Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From hbrock at redhat.com Thu Mar 1 15:28:16 2007 From: hbrock at redhat.com (Hugh Brock) Date: Thu, 01 Mar 2007 10:28:16 -0500 Subject: [et-mgmt-tools] [PATCH] waiting VNC port number take2 In-Reply-To: <20070301154622.B2DC.FUKUTA.SAORI@jp.fujitsu.com> References: <20070223105543.DAE7.FUKUTA.SAORI@jp.fujitsu.com> <20070223120004.GA14993@redhat.com> <20070301154622.B2DC.FUKUTA.SAORI@jp.fujitsu.com> Message-ID: <45E6F110.80705@redhat.com> Saori Fukuta wrote: > Hi, > > On Fri, 23 Feb 2007 12:00:05 +0000 "Daniel P. Berrange" wrote: > >>>> - Once the above fix in libvirt is done, we can make virt-install do >>>> a loop like the one you have below, but instead of looking in xenstore >>>> it will look at the port in the XML - waiting for it to change from '-1' >>>> to an actual port number >>> You are right, I will consider again as you pointed out. >> Excellant. > > I made a patch for current libvirt(revision 1.420) that fixed following: > revision 1.417 > date: 2007/02/27 15:22:13; author: berrange; state: Exp; lines:+7 -0 > Don't hardcode port=5900+domid for new xend > > Signed-off-by: Saori Fukuta > > Thanks, > Saori Fukuta. > > Index: virt-install > =================================================================== > diff -r 1776ef836bf6 virt-install > --- a/virt-install Wed Feb 28 07:47:31 2007 -0500 > +++ b/virt-install Thu Mar 01 23:26:17 2007 +0900 > @@ -313,11 +313,19 @@ def get_xml_string(dom, path): > return None > > def vnc_console(dom): > - import time; time.sleep(2) # FIXME: ugh. > - vncport = get_xml_string(dom, > - "/domain/devices/graphics[@type='vnc']/@port") > - if vncport == None: > - vncport = 5900 + dom.ID() > + import time; > + num = 0 > + while num < ( 40 / 0.25 ): # 40 seconds, .25 second sleeps > + vncport = get_xml_string(dom, > + "/domain/devices/graphics[@type='vnc']/@port") > + if vncport == '-1': > + num += 1 > + time.sleep(0.25) > + else: > + break > + if vncport == '-1' or vncport is None: > + print >> sys.stderr, "Unable to connect to graphical console; vncport number is not found." > + return None > vncport = int(vncport) > vnchost = "localhost" > if not os.path.exists("/usr/bin/vncviewer"): > Thank you Saori-san. Patch applied: changeset: 114:5a1190f54c80 tag: tip user: "Hugh O. Brock " date: Thu Mar 01 10:24:34 2007 -0500 summary: Make the vnc port number acquisition sane. Specifically, sleep in .25 second increments for 40 seconds while waiting for the guest's vnc server to wake up. Patch from Saori Fukuta . --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From cobbler at msquared.id.au Fri Mar 2 05:45:54 2007 From: cobbler at msquared.id.au (Msquared) Date: Fri, 2 Mar 2007 14:45:54 +0900 Subject: [et-mgmt-tools] Released -- Cobbler 0.4.3 - kickstart path generation bug on import In-Reply-To: <45E60641.5060209@redhat.com> References: <45E60641.5060209@redhat.com> Message-ID: <20070302054554.GA519@sliderule.msquared.com.au> On Wed, Feb 28, 2007 at 05:46:25PM -0500, Michael DeHaan wrote: > Cobbler 0.4.3 Bug in import, producing kickstart tree. eg: cobbler --import --mirror=root at localhost:/media/cdrom/ --mirror-name=fedora6 produces (among the rest of the output) *** KICKSTART TREE = tree=http://134.115.xxx.xxx//ks_mirror/fedora6 which should probably read *** KICKSTART TREE = tree=http://134.115.xxx.xxx/cobbler/ks_mirror/fedora6 Regards, Msquared... From cobbler at msquared.id.au Fri Mar 2 06:49:54 2007 From: cobbler at msquared.id.au (Msquared) Date: Fri, 2 Mar 2007 15:49:54 +0900 Subject: [et-mgmt-tools] Released -- Cobbler 0.4.3 - kickstart selection bug? In-Reply-To: <45E60641.5060209@redhat.com> References: <45E60641.5060209@redhat.com> Message-ID: <20070302064954.GB519@sliderule.msquared.com.au> On Wed, Feb 28, 2007 at 05:46:25PM -0500, Michael DeHaan wrote: > Cobbler 0.4.3 When I ran an import of a DVD with cobbler --import --mirror=root at localhost:/media/cdrom/ --mirror-name=fedora6 the profiles (one for pxeboot, one for xen) both had virt_paravirt: True set, and they both had /etc/kickstart_fc5.ks selected, instead of kickstart_fc6.ks or kickstart_fc6_DomU.ks selected. Regards, Msquared... From cobbler at msquared.id.au Fri Mar 2 07:13:30 2007 From: cobbler at msquared.id.au (Msquared) Date: Fri, 2 Mar 2007 16:13:30 +0900 Subject: [et-mgmt-tools] ks_meta in profile vs distro Message-ID: <20070302071330.GN13894@sliderule.msquared.com.au> Hi! I have a question about how dictionaries are handled in distro entries and profile entries, and a suggestion about the default location of the tree key in the distro and profile entries that are created with cobbler import. Handling Dictionaries In Distros and Profiles I notice that the ks_meta dictionary exists in both the profile and the distro: does this mean that the dictionaries are merged before being used? If a dictionary key is in both, will the entry in the profile override the entry in the distro? Default Location For tree in ks_meta Dictionary When I run cobbler import, the ks_meta tree entry ends up in the profile that is created, but can that be added to the distro instead? It seems to me that the tree is associated with the distro, so the distro would be the most sensible place for the tree key of the ks_meta dictionary to be set. I realise that I could move the tree entry from the profile to the distro (assuming they are merged), but I would prefer fewer manual steps. While it makes most sense to set the tree in the distro, I imagine that someone else might prefer to keep it in the profile, so perhaps there could be a command-line option for this? Regards, Msquared... From cobbler at msquared.id.au Fri Mar 2 07:16:03 2007 From: cobbler at msquared.id.au (Msquared) Date: Fri, 2 Mar 2007 16:16:03 +0900 Subject: [et-mgmt-tools] kickstart file specified in distro (rather than profile)? Message-ID: <20070302071603.GO13894@sliderule.msquared.com.au> Can I specify a kickstart entry in the distro? This would be the default, which I could override for each profile I create. I ask because I notice the kickstart option to cobbler profile add is optional. Perhaps cobbler import could have an option that allows the kickstart entry to be written to the distro it creates, rather than to the profile it creates? Regards, Msquared... From cobbler at msquared.id.au Fri Mar 2 07:19:12 2007 From: cobbler at msquared.id.au (Msquared) Date: Fri, 2 Mar 2007 16:19:12 +0900 Subject: [et-mgmt-tools] profile 'template' mechanism? Message-ID: <20070302071912.GP13894@sliderule.msquared.com.au> Would a template mechanism for profiles be feasible? The simplest thing I could think would be something like this: cobbler profile add --name=customprofile2 --based-on=standardprofile --virt-ram=128 This would create a profile called customprofile2 that is a copy of standardprofile, but with a different virtual ram size. Thoughts? Regards, Msquared... From cobbler at msquared.id.au Fri Mar 2 07:29:47 2007 From: cobbler at msquared.id.au (Msquared) Date: Fri, 2 Mar 2007 16:29:47 +0900 Subject: [et-mgmt-tools] overwriting a profile - desired? Message-ID: <20070302072947.GC519@sliderule.msquared.com.au> I notice that if I issue the same 'cobbler profile add' command a second time, it overwrites the original entry it created, without warning or confirmation. While I can see how this might be useful while you're creating a profile (oops, forgot to add an option, use command history to recall and edit the command), I think this could otherwise be dangerous. Perhaps cobbler should generate an error unless you issue '--force' on the command line? Regards, Msquared... From mdehaan at redhat.com Fri Mar 2 15:39:24 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 02 Mar 2007 10:39:24 -0500 Subject: [et-mgmt-tools] Released -- Cobbler 0.4.3 - kickstart selection bug? In-Reply-To: <20070302064954.GB519@sliderule.msquared.com.au> References: <45E60641.5060209@redhat.com> <20070302064954.GB519@sliderule.msquared.com.au> Message-ID: <45E8452C.8040702@redhat.com> Msquared wrote: > On Wed, Feb 28, 2007 at 05:46:25PM -0500, Michael DeHaan wrote: > > >> Cobbler 0.4.3 >> > > When I ran an import of a DVD with > > cobbler --import --mirror=root at localhost:/media/cdrom/ --mirror-name=fedora6 > > the profiles (one for pxeboot, one for xen) both had virt_paravirt: True > set, and they both had /etc/kickstart_fc5.ks selected, instead of > kickstart_fc6.ks or kickstart_fc6_DomU.ks selected. > > Regards, Msquared... > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > Good catch. This is because your creative use of the rsync import feature isn't going against a full tree as it was originally intended. We rely on the tree looking something like a public mirror (with directories like FC6) in order to choose the correct kickstart. Currently, you can edit /var/lib/cobbler/profiles and then run "cobbler sync" to make corrections as you would want. It's true that the automatic kickstart mapping isn't programmed to ever pick "kickstart_fc6_domU.ks", though it probably should learn how to do this. On the other hand, I'm not entirely sure, as that kickstart is an example of a minimal domU image, which may not be appropriate for everyone. Import should basically be viewed as a way to start building a cobbler configuration, though a user is probably going to want to modify the assigned kickstarts after the fact as well as create some additional profiles. --MPD From mdehaan at redhat.com Fri Mar 2 15:41:20 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 02 Mar 2007 10:41:20 -0500 Subject: [et-mgmt-tools] kickstart file specified in distro (rather than profile)? In-Reply-To: <20070302071603.GO13894@sliderule.msquared.com.au> References: <20070302071603.GO13894@sliderule.msquared.com.au> Message-ID: <45E845A0.8090207@redhat.com> Msquared wrote: > Can I specify a kickstart entry in the distro? This would be the default, > which I could override for each profile I create. > > I ask because I notice the kickstart option to cobbler profile add is > optional. > > Perhaps cobbler import could have an option that allows the kickstart > entry to be written to the distro it creates, rather than to the profile > it creates? > > Regards, Msquared... > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > It belongs in the profile. A kickstart explains /how/ a distribution should be installed, not /what/ distribution should be installed. Multiple profiles can use a single distro. For instance, a 'fc6webserver' and a 'fc6dbserver' could both use a 'fc6x86_64' distro. From mdehaan at redhat.com Fri Mar 2 15:42:08 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 02 Mar 2007 10:42:08 -0500 Subject: [et-mgmt-tools] profile 'template' mechanism? In-Reply-To: <20070302071912.GP13894@sliderule.msquared.com.au> References: <20070302071912.GP13894@sliderule.msquared.com.au> Message-ID: <45E845D0.1090502@redhat.com> Msquared wrote: > Would a template mechanism for profiles be feasible? > > The simplest thing I could think would be something like this: > > cobbler profile add --name=customprofile2 --based-on=standardprofile --virt-ram=128 > > This would create a profile called customprofile2 that is a copy of > standardprofile, but with a different virtual ram size. > > Thoughts? > > Regards, Msquared... > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > I actually thought about this yesterday... someone could send in a patch, maybe :) From mdehaan at redhat.com Fri Mar 2 15:43:20 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 02 Mar 2007 10:43:20 -0500 Subject: [et-mgmt-tools] overwriting a profile - desired? In-Reply-To: <20070302072947.GC519@sliderule.msquared.com.au> References: <20070302072947.GC519@sliderule.msquared.com.au> Message-ID: <45E84618.5010705@redhat.com> Msquared wrote: > I notice that if I issue the same 'cobbler profile add' command a second > time, it overwrites the original entry it created, without warning or > confirmation. > > While I can see how this might be useful while you're creating a profile > (oops, forgot to add an option, use command history to recall and edit the > command), I think this could otherwise be dangerous. > > Perhaps cobbler should generate an error unless you issue '--force' on the > command line? > > Regards, Msquared... > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > This suggestion came up a while back. This would be a really good patch to send in also :) Ideally "add" would generate warnings, and "edit" would not. Right now, "add" and "edit" mean the same thing. --Michael From schruefer at scai.fhg.de Fri Mar 2 09:40:39 2007 From: schruefer at scai.fhg.de (Eberhard Schruefer) Date: Fri, 02 Mar 2007 10:40:39 +0100 Subject: [et-mgmt-tools] Is --repos working? Message-ID: <45E7F117.6030302@scai.fhg.de> Hi, I'm trying to add repositories to a profile but I'm ending up with a traceback. This happens with cobbler 0.4.3 on a FC6 server. Am I doing something wrong or is this a bug? Thanks Eberhard # cobbler profile add --name=FC6workstation --distro=var_www_cobbler_ks_mirror_FC6_i386_os_images_pxeboot --kickstart=/export/INSTALL/Lin/FC6/ks/ks.cfg --repos="FC6_updates_i386 FC6_extras_i386" Traceback (most recent call last): File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 505, in main BootCLI(sys.argv).run() File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 102, in run self.curry_args(self.args[1:], self.commands['toplevel']) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 421, in curry_args commands[args[0]](args[1:]) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 475, in profile return self.curry_args(args, self.commands['profile']) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 421, in curry_args commands[args[0]](args[1:]) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 353, in profile_edit return self.apply_args(args,commands,on_ok) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 406, in apply_args input_routines[key](value) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 350, in '--repos' : lambda(a) : profile.set_repos(a) File "/usr/lib/python2.4/site-packages/cobbler/item_profile.py", line 93, in set_repos repolist.remove('') ValueError: list.remove(x): x not in list From mdehaan at redhat.com Fri Mar 2 15:51:27 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 02 Mar 2007 10:51:27 -0500 Subject: [et-mgmt-tools] Released -- Cobbler 0.4.3 - kickstart path generation bug on import In-Reply-To: <20070302054554.GA519@sliderule.msquared.com.au> References: <45E60641.5060209@redhat.com> <20070302054554.GA519@sliderule.msquared.com.au> Message-ID: <45E847FF.4040507@redhat.com> Msquared wrote: > On Wed, Feb 28, 2007 at 05:46:25PM -0500, Michael DeHaan wrote: > > >> Cobbler 0.4.3 >> > > Bug in import, producing kickstart tree. eg: > > cobbler --import --mirror=root at localhost:/media/cdrom/ --mirror-name=fedora6 > > produces (among the rest of the output) > > *** KICKSTART TREE = tree=http://134.115.xxx.xxx//ks_mirror/fedora6 > > which should probably read > > *** KICKSTART TREE = tree=http://134.115.xxx.xxx/cobbler/ks_mirror/fedora6 > > Regards, Msquared... > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > The fix is ... diff -r a988cddaa21f176464480efa8b99370ed89b8d76 cobbler/action_import.py --- a/cobbler/action_import.py Wed Feb 28 18:29:30 2007 -0500 +++ b/cobbler/action_import.py Fri Mar 02 10:49:40 2007 -0500 @@ -185,7 +185,7 @@ class Importer: tokens = tokens[:-2] base = "/".join(tokens) base = base.replace(self.settings.webdir,"") - tree = "tree=http://%s/%s" % (self.settings.server, base) + tree = "tree=http://%s/cobbler_track/%s" % (self.settings.server, base) print "*** KICKSTART TREE = %s" % tree profile.set_ksmeta(tree) self.serialize_counter = self.serialize_counter + 1 Thanks! The "cobbler_track" URL is used here instead of just "cobbler" because some of the kickstart tracking bits know how to pay attention to "cobbler_track" URLs. They don't read log regular cobbler URL's to cut down on noise. --Michael From mdehaan at redhat.com Fri Mar 2 15:56:25 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 02 Mar 2007 10:56:25 -0500 Subject: [et-mgmt-tools] ks_meta in profile vs distro In-Reply-To: <20070302071330.GN13894@sliderule.msquared.com.au> References: <20070302071330.GN13894@sliderule.msquared.com.au> Message-ID: <45E84929.7040906@redhat.com> Msquared wrote: > Hi! > > I have a question about how dictionaries are handled in distro entries and > profile entries, and a suggestion about the default location of the tree > key in the distro and profile entries that are created with cobbler > import. > > > Handling Dictionaries In Distros and Profiles > > I notice that the ks_meta dictionary exists in both the profile and the > distro: does this mean that the dictionaries are merged before being used? > If a dictionary key is in both, will the entry in the profile override the > entry in the distro? > It will. Systems override profiles, and profiles override distros. > > Default Location For tree in ks_meta Dictionary > > When I run cobbler import, the ks_meta tree entry ends up in the profile > that is created, but can that be added to the distro instead? > > It seems to me that the tree is associated with the distro, so the distro > would be the most sensible place for the tree key of the ks_meta > dictionary to be set. > This is correct -- in the end (functionally speaking), it really doesn't make a difference other than in the eyes of the person looking at the configuration files. It's true that the kickstart tree location is associated with the distribution, but since the particular profile created is tied to the distribution as well, it all works out -- it's the same result in terms of a output file from the template. It would be a good change to make technically speaking, but would not affect results. I would take a patch for it though. > I realise that I could move the tree entry from the profile to the distro > (assuming they are merged), but I would prefer fewer manual steps. > > While it makes most sense to set the tree in the distro, I imagine that > someone else might prefer to keep it in the profile, so perhaps there > could be a command-line option for this? > An option shouldn't be needed... one way or the other is fine. Cobbler has a bit too many options already :) > > Regards, Msquared... > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > Thanks, Michael From tforeman at ibsys.com Fri Mar 2 18:16:17 2007 From: tforeman at ibsys.com (Foreman, Tim) Date: Fri, 2 Mar 2007 12:16:17 -0600 Subject: [et-mgmt-tools] Koan and Xen question Message-ID: <514B41B4EB0E834EAFB17FA73E263B7009B2EC@exch001.ibsys.com> I'm not sure if this is a Koan issue, a Xen issue or an "I'm stupid" issue, but I'm going to start with Koan. :-) I've built a Fedora Core 6 box with Xen and Koan installed. I have another box with Cobbler on it and I'm trying to kickstart a Xen guest domain with Koan. I'm trying to build a RH EL 4 U4 guest (which might be part of my problem.) I have a profile setup using the default.ks file. When I run: 'koan --virt --server=kickstart.mtc.ibsys.com --profile=rh4u4_xen_domU' I get the following output: - processing profile: rh4u4_xen_domU - fetching configuration for profile: rh4u4_xen_domU - url=http://kickstart.mtc.ibsys.com/cobbler/profiles/rh4u4_xen_domU - {'kickstart': 'http://kickstart.mtc.ibsys.com/cobbler_track/kickstarts/rh4u4_xen_domU/ks.c fg', 'name': 'rh4u4_xen_domU', 'virt_ram': 512, 'repos': '', 'kernel_options': 'ksdevice=eth0 lang= vga=788 syslog=kickstart.mtc.ibsys.com:25150 devfs=nomount ramdisk_size=16438 append ', 'virt_name': 'virtguest', 'virt_file_size': 5, 'distro': 'rhel4u4', 'virt_paravirt': 'True', 'ks_meta': ''} - fetching configuration for distro: rhel4u4 - url=http://kickstart.mtc.ibsys.com/cobbler/distros/rhel4u4 - {'kernel': '/var/www/html/kickstart/rhel4u4/i386/isolinux/vmlinuz', 'ks_meta': '', 'breed': 'redhat', 'kernel_options': 'ksdevice=eth0 lang= vga=788 syslog=kickstart.mtc.ibsys.com:25150 devfs=nomount ramdisk_size=16438 append ', 'initrd': '/var/www/html/kickstart/rhel4u4/i386/isolinux/initrd.img', 'arch': 'x86', 'name': 'rhel4u4'} - downloading initrd initrd.img to /tmp/initrd.img - url=http://kickstart.mtc.ibsys.com/cobbler/images/rhel4u4/initrd.img - downloading kernel vmlinuz to /tmp/vmlinuz - url=http://kickstart.mtc.ibsys.com/cobbler/images/rhel4u4/vmlinuz - kernel saved = /tmp/vmlinuz - initrd saved = /tmp/initrd.img libvir: Xen Daemon error : POST operation failed: (xend.err "Error creating domain: (22, 'Invalid argument')") Failed to create domain virtguest Traceback (most recent call last): File "/usr/lib/python2.3/site-packages/koan/app.py", line 97, in main k.run() File "/usr/lib/python2.3/site-packages/koan/app.py", line 155, in run self.do_virt() File "/usr/lib/python2.3/site-packages/koan/app.py", line 300, in do_virt return self.do_net_install("/tmp",after_download) File "/usr/lib/python2.3/site-packages/koan/app.py", line 256, in do_net_install after_download(self, distro_data, profile_data) File "/usr/lib/python2.3/site-packages/koan/app.py", line 299, in after_download self.do_virt_net_install(profile_data, distro_data) File "/usr/lib/python2.3/site-packages/koan/app.py", line 599, in do_virt_net_install extra=kextra File "/usr/lib/python2.3/site-packages/koan/virtcreate.py", line 201, in start_paravirt_install dom = conn.createLinux(cfgxml, 0) File "/usr/lib/python2.4/site-packages/libvirt.py", line 249, in createLinux if ret is None:raise libvirtError('virDomainCreateLinux() failed') libvirtError: virDomainCreateLinux() failed So, is this a Koan issue? If so, what's happening? Is it a Cobbler issue? Did I not put something in my profile? This is my profile: profile : rh4u4_xen_domU distro : rhel4u4 kickstart : /etc/cobbler/default.ks kernel options : {} ks metadata : {} virt name : virtguest virt file size : 5 virt ram : 512 virt paravirt : True repos : If it's a Xen issue, tell me and I'll try the Xen users list (unless you have the solution and want to send it to me. :-) If it's an "I'm stupid" issue, please tell me what I'm doing wrong. :-) I've been using Cobbler for a while (and love it) but I'm new to Xen and Koan. Thanks! -- Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com -- The Onion: Have you decided what you want to be when you grow up? Berkeley Breathed: Dad. The rest is frosting. From David.Mackintosh at xdroop.com Fri Mar 2 19:51:25 2007 From: David.Mackintosh at xdroop.com (David Mackintosh) Date: Fri, 2 Mar 2007 14:51:25 -0500 Subject: [et-mgmt-tools] distro add error. Message-ID: <20070302195125.GA29591@xdroop.com> I'm probably bending the envelope again but can't figure out how. What is cobbler trying to tell me I've done wrong? [root at helmsdeep ~]# rpm -qa | grep cobb cobbler-0.4.2-0 [root at helmsdeep ~]# rpm -qa | grep cheet python23-cheetah-0.9.18-1.el4.pyv python-cheetah-0.9.18-1.el4.pyv Red Hat Enterprise Linux AS release 4 (Nahant Update 2) [root at helmsdeep ~]# uname -a Linux helmsdeep 2.6.9-22.EL #1 Mon Sep 19 17:49:49 EDT 2005 x86_64 x86_64 x86_64 GNU/Linux [root at helmsdeep ~]# ls "/mnt/kickstart/CentOS-4.3-i386/images/pxeboot/vmlinuz" /mnt/kickstart/CentOS-4.3-i386/images/pxeboot/vmlinuz [root at helmsdeep ~]# cat /etc/redhat-release [root at helmsdeep ~]# cobbler distro add --name="CentOS-4.3-i386" --kernel="/mnt/kickstart/CentOS-4.3-i386/images/pxeboot/vmlinuz" --initrd="/mnt/kickstart/CentOS-4.3-i386/images/pxeboot/initrd.img" Traceback (most recent call last): File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 505, in main BootCLI(sys.argv).run() File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 102, in run self.curry_args(self.args[1:], self.commands['toplevel']) File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 421, in curry_args commands[args[0]](args[1:]) File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 469, in distro return self.curry_args(args, self.commands['distro']) File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 421, in curry_args commands[args[0]](args[1:]) File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 386, in distro_edit return self.apply_args(args,commands,on_ok) File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 409, in apply_args on_ok() File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 385, in on_ok = lambda: self.api.distros().add(distro, with_copy=self.api.sync_flag) File "/usr/lib/python2.3/site-packages/cobbler/collection.py", line 97, in add lite_sync.add_single_distro(ref.name) File "/usr/lib/python2.3/site-packages/cobbler/action_litesync.py", line 60, in add_single_distro self.sync.write_distro_file(distro) File "/usr/lib/python2.3/site-packages/cobbler/action_sync.py", line 733, in write_distro_file clone.kernel_options = self.hash_to_string(clone.kernel_options) File "/usr/lib/python2.3/site-packages/cobbler/action_sync.py", line 885, in hash_to_string buffer = buffer + key + "=" + value + " " TypeError: cannot concatenate 'str' and 'int' objects -- /\oo/\ / /()\ \ David Mackintosh | dave at xdroop.com | http://www.xdroop.com -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From mdehaan at redhat.com Fri Mar 2 20:30:53 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 02 Mar 2007 15:30:53 -0500 Subject: [et-mgmt-tools] Koan and Xen question In-Reply-To: <514B41B4EB0E834EAFB17FA73E263B7009B2EC@exch001.ibsys.com> References: <514B41B4EB0E834EAFB17FA73E263B7009B2EC@exch001.ibsys.com> Message-ID: <45E8897D.5060604@redhat.com> Foreman, Tim wrote: > I'm not sure if this is a Koan issue, a Xen issue or an "I'm stupid" > issue, but I'm going to start with Koan. :-) > > I've built a Fedora Core 6 box with Xen and Koan installed. > > I have another box with Cobbler on it and I'm trying to kickstart a > Xen guest domain with Koan. > > I'm trying to build a RH EL 4 U4 guest (which might be part of my > problem.) > > I have a profile setup using the default.ks file. > > When I run: > 'koan --virt --server=kickstart.mtc.ibsys.com --profile=rh4u4_xen_domU' > > I get the following output: > > - processing profile: rh4u4_xen_domU > - fetching configuration for profile: rh4u4_xen_domU > - url=http://kickstart.mtc.ibsys.com/cobbler/profiles/rh4u4_xen_domU > - {'kickstart': > 'http://kickstart.mtc.ibsys.com/cobbler_track/kickstarts/rh4u4_xen_domU/ks.c > fg', 'name': 'rh4u4_xen_domU', 'virt_ram': 512, 'repos': '', > 'kernel_options': 'ksdevice=eth0 lang= vga=788 > syslog=kickstart.mtc.ibsys.com:25150 devfs=nomount ramdisk_size=16438 append > ', 'virt_name': 'virtguest', 'virt_file_size': 5, 'distro': 'rhel4u4', > 'virt_paravirt': 'True', 'ks_meta': ''} > - fetching configuration for distro: rhel4u4 > - url=http://kickstart.mtc.ibsys.com/cobbler/distros/rhel4u4 > - {'kernel': '/var/www/html/kickstart/rhel4u4/i386/isolinux/vmlinuz', > 'ks_meta': '', 'breed': 'redhat', 'kernel_options': 'ksdevice=eth0 lang= > vga=788 syslog=kickstart.mtc.ibsys.com:25150 devfs=nomount > ramdisk_size=16438 append ', 'initrd': > '/var/www/html/kickstart/rhel4u4/i386/isolinux/initrd.img', 'arch': 'x86', > 'name': 'rhel4u4'} > - downloading initrd initrd.img to /tmp/initrd.img > - url=http://kickstart.mtc.ibsys.com/cobbler/images/rhel4u4/initrd.img > - downloading kernel vmlinuz to /tmp/vmlinuz > - url=http://kickstart.mtc.ibsys.com/cobbler/images/rhel4u4/vmlinuz > - kernel saved = /tmp/vmlinuz > - initrd saved = /tmp/initrd.img > libvir: Xen Daemon error : POST operation failed: (xend.err "Error creating > domain: (22, 'Invalid argument')") > Failed to create domain virtguest > Traceback (most recent call last): > File "/usr/lib/python2.3/site-packages/koan/app.py", line 97, in main > k.run() > File "/usr/lib/python2.3/site-packages/koan/app.py", line 155, in run > self.do_virt() > File "/usr/lib/python2.3/site-packages/koan/app.py", line 300, in do_virt > return self.do_net_install("/tmp",after_download) > File "/usr/lib/python2.3/site-packages/koan/app.py", line 256, in > do_net_install > after_download(self, distro_data, profile_data) > File "/usr/lib/python2.3/site-packages/koan/app.py", line 299, in > after_download > self.do_virt_net_install(profile_data, distro_data) > File "/usr/lib/python2.3/site-packages/koan/app.py", line 599, in > do_virt_net_install > extra=kextra > File "/usr/lib/python2.3/site-packages/koan/virtcreate.py", line 201, in > start_paravirt_install > dom = conn.createLinux(cfgxml, 0) > File "/usr/lib/python2.4/site-packages/libvirt.py", line 249, in > createLinux > if ret is None:raise libvirtError('virDomainCreateLinux() failed') > libvirtError: virDomainCreateLinux() failed > > So, is this a Koan issue? If so, what's happening? Is it a Cobbler issue? > Did I not put something in my profile? > > This is my profile: > > profile : rh4u4_xen_domU > distro : rhel4u4 > kickstart : /etc/cobbler/default.ks > kernel options : {} > ks metadata : {} > virt name : virtguest > virt file size : 5 > virt ram : 512 > virt paravirt : True > repos : > > If it's a Xen issue, tell me and I'll try the Xen users list (unless you > have > the solution and want to send it to me. :-) > > If it's an "I'm stupid" issue, please tell me what I'm doing wrong. :-) > I've been using Cobbler for a while (and love it) but I'm new to Xen > and Koan. > > Thanks! > -- > Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com > (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com > -- > The Onion: Have you decided what you want to be when you grow up? > Berkeley Breathed: Dad. The rest is frosting. > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > The distro needs to reference a domU kernel. You're going the xen kernel/initrd from the RHEL4u5 public beta, or the Fedora xen images. I've been told that the RHEL4u5 beta went live on Feb 26th through RHN. --Michael From mdehaan at redhat.com Fri Mar 2 20:34:43 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 02 Mar 2007 15:34:43 -0500 Subject: [et-mgmt-tools] distro add error. In-Reply-To: <20070302195125.GA29591@xdroop.com> References: <20070302195125.GA29591@xdroop.com> Message-ID: <45E88A63.5030804@redhat.com> David Mackintosh wrote: > I'm probably bending the envelope again but can't figure out how. > What is cobbler trying to tell me I've done wrong? > > [root at helmsdeep ~]# rpm -qa | grep cobb > cobbler-0.4.2-0 > [root at helmsdeep ~]# rpm -qa | grep cheet > python23-cheetah-0.9.18-1.el4.pyv > python-cheetah-0.9.18-1.el4.pyv > Red Hat Enterprise Linux AS release 4 (Nahant Update 2) > [root at helmsdeep ~]# uname -a > Linux helmsdeep 2.6.9-22.EL #1 Mon Sep 19 17:49:49 EDT 2005 x86_64 x86_64 x86_64 GNU/Linux > [root at helmsdeep ~]# ls "/mnt/kickstart/CentOS-4.3-i386/images/pxeboot/vmlinuz" /mnt/kickstart/CentOS-4.3-i386/images/pxeboot/vmlinuz > [root at helmsdeep ~]# cat /etc/redhat-release > [root at helmsdeep ~]# cobbler distro add --name="CentOS-4.3-i386" --kernel="/mnt/kickstart/CentOS-4.3-i386/images/pxeboot/vmlinuz" --initrd="/mnt/kickstart/CentOS-4.3-i386/images/pxeboot/initrd.img" > Traceback (most recent call last): > File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 505, in main > BootCLI(sys.argv).run() > File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 102, in run > self.curry_args(self.args[1:], self.commands['toplevel']) > File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 421, in curry_args > commands[args[0]](args[1:]) > File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 469, in distro > return self.curry_args(args, self.commands['distro']) > File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 421, in curry_args > commands[args[0]](args[1:]) > File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 386, in distro_edit > return self.apply_args(args,commands,on_ok) > File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 409, in apply_args > on_ok() > File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 385, in > on_ok = lambda: self.api.distros().add(distro, with_copy=self.api.sync_flag) > File "/usr/lib/python2.3/site-packages/cobbler/collection.py", line 97, in add > lite_sync.add_single_distro(ref.name) > File "/usr/lib/python2.3/site-packages/cobbler/action_litesync.py", line 60, in add_single_distro > self.sync.write_distro_file(distro) > File "/usr/lib/python2.3/site-packages/cobbler/action_sync.py", line 733, in write_distro_file > clone.kernel_options = self.hash_to_string(clone.kernel_options) > File "/usr/lib/python2.3/site-packages/cobbler/action_sync.py", line 885, in hash_to_string > buffer = buffer + key + "=" + value + " " > TypeError: cannot concatenate 'str' and 'int' objects > > > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools It's a bug. There's a patch regarding this a bit further up on the mailing list, or you can just upgrade to 0.4.3 and this goes away. David Lutterkort runs a RHEL4 yum repo that includes cobbler and koan here -- http://people.redhat.com/~dlutter/yum/rhel4/ --Michael From John.Sanabria at ece.uprm.edu Fri Mar 2 20:37:57 2007 From: John.Sanabria at ece.uprm.edu (John Sanabria) Date: Fri, 2 Mar 2007 16:37:57 -0400 (AST) Subject: [et-mgmt-tools] cobbler enchant problem... Message-ID: <38130.136.145.116.114.1172867877.squirrel@ece.uprm.edu> Hi, i have a functional "cobbler server". Its ip number is 136.145.116.114. With that cobbler server i can create virtual boxes inside 136...114 machine. Now, i have another machine 136.145.116.88 and i had install cobbler and koan on it, the /var/lib/cobbler/settings file, i had modify as follows: next_server: '136.145.116.114' server: '136.145.116.114' Now, when i log as root in 136...114 and run cobbler enchant --virt=yes --address=136.145.116.88 --profile="xentemplate-profile" enchant failed (koan is missing) any clues about it? is necessary to remote system configure any cobbler/koan tools? --- additional info --- [root at reboot ~]# cobbler report distro : xentemplate kernel : /root/vmlinuz initrd : /root/initrd.img kernel options : {} architecture : x86 ks metadata : {} breed : redhat profile : xentemplate-profile distro : xentemplate kickstart : /etc/cobbler/grid.cfg kernel options : {} ks metadata : {} virt name : xentemplate virt file size : 6 virt ram : 256 virt paravirt : True repos : http://ece.uprm.edu/~s047267 http://del.icio.us/josanabr http://blog-grid.blogspot.com From mdehaan at redhat.com Fri Mar 2 20:46:49 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 02 Mar 2007 15:46:49 -0500 Subject: [et-mgmt-tools] cobbler enchant problem... In-Reply-To: <38130.136.145.116.114.1172867877.squirrel@ece.uprm.edu> References: <38130.136.145.116.114.1172867877.squirrel@ece.uprm.edu> Message-ID: <45E88D39.5050208@redhat.com> John Sanabria wrote: > Hi, > > i have a functional "cobbler server". Its ip number is 136.145.116.114. > With that cobbler server i can create virtual boxes inside 136...114 > machine. Now, i have another machine 136.145.116.88 and i had install > cobbler and koan on it, the /var/lib/cobbler/settings file, i had modify > as follows: > > next_server: '136.145.116.114' > server: '136.145.116.114' > > Now, when i log as root in 136...114 and run > > cobbler enchant --virt=yes --address=136.145.116.88 > --profile="xentemplate-profile" > enchant failed (koan is missing) > > any clues about it? is necessary to remote system configure any > cobbler/koan tools? > This one's pretty easy. Add the full path to the RPM of 'koan' to koan_path in /var/lib/cobbler/settings. You can get the latest noarch RPM from et.redhat.com. (Don't put it in your cobbler web directory, just anywhere on the filesystem is fine. /var/lib/cobbler/koan*.rpm would be a good choice). Then run "cobbler sync" to make sure cobbler copies the koan RPM to the right place. Looks like I need to improve that error message some to make that clearer. --Michael From John.Sanabria at ece.uprm.edu Fri Mar 2 21:02:59 2007 From: John.Sanabria at ece.uprm.edu (John Sanabria) Date: Fri, 2 Mar 2007 17:02:59 -0400 (AST) Subject: [et-mgmt-tools] cobbler enchant problem... In-Reply-To: <45E88D39.5050208@redhat.com> References: <38130.136.145.116.114.1172867877.squirrel@ece.uprm.edu> <45E88D39.5050208@redhat.com> Message-ID: <41084.136.145.116.114.1172869379.squirrel@ece.uprm.edu> Many thanks, now is working. happy weekend :-D > John Sanabria wrote: >> Hi, >> >> i have a functional "cobbler server". Its ip number is 136.145.116.114. >> With that cobbler server i can create virtual boxes inside 136...114 >> machine. Now, i have another machine 136.145.116.88 and i had install >> cobbler and koan on it, the /var/lib/cobbler/settings file, i had modify >> as follows: >> >> next_server: '136.145.116.114' >> server: '136.145.116.114' >> >> Now, when i log as root in 136...114 and run >> >> cobbler enchant --virt=yes --address=136.145.116.88 >> --profile="xentemplate-profile" >> enchant failed (koan is missing) >> >> any clues about it? is necessary to remote system configure any >> cobbler/koan tools? >> > This one's pretty easy. Add the full path to the RPM of 'koan' to > koan_path in /var/lib/cobbler/settings. You can get the latest noarch > RPM from et.redhat.com. (Don't put it in your cobbler web directory, > just anywhere on the filesystem is fine. /var/lib/cobbler/koan*.rpm > would be a good choice). > > Then run "cobbler sync" to make sure cobbler copies the koan RPM to the > right place. > > Looks like I need to improve that error message some to make that clearer. > > --Michael > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > http://ece.uprm.edu/~s047267 http://del.icio.us/josanabr http://blog-grid.blogspot.com From ccoffing at novell.com Sat Mar 3 01:13:01 2007 From: ccoffing at novell.com (Charles Coffing) Date: Fri, 02 Mar 2007 18:13:01 -0700 Subject: [et-mgmt-tools] [PATCH] make F10 pass-through to guest Message-ID: <45E86882.D169.003C.0@novell.com> By default, F10 is bound to virt-manager's menu bar, so the guest can never get F10. The attached patch disables this accelerator for the virt-manager app when the keyboard is grabbed, and restores it when the keyboard is released. -------------- next part -------------- A non-text attachment was scrubbed... Name: virtman-f10.diff Type: application/octet-stream Size: 924 bytes Desc: not available URL: From David.Mackintosh at xdroop.com Sat Mar 3 02:18:47 2007 From: David.Mackintosh at xdroop.com (David Mackintosh) Date: Fri, 2 Mar 2007 21:18:47 -0500 Subject: [et-mgmt-tools] distro add error. In-Reply-To: <45E88A63.5030804@redhat.com> References: <20070302195125.GA29591@xdroop.com> <45E88A63.5030804@redhat.com> Message-ID: <20070303021847.GB29591@xdroop.com> On Fri, Mar 02, 2007 at 03:34:43PM -0500, Michael DeHaan wrote: > David Mackintosh wrote: > >I'm probably bending the envelope again but can't figure out how. > >What is cobbler trying to tell me I've done wrong? > It's a bug. There's a patch regarding this a bit further up on the > mailing list, or you can just upgrade to 0.4.3 and this goes away. Yes, upgrading to 0.4.3 corrected the problem. Thanks. -- /\oo/\ / /()\ \ David Mackintosh | dave at xdroop.com | http://www.xdroop.com -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 189 bytes Desc: not available URL: From jeevanullas at gmail.com Mon Mar 5 04:42:36 2007 From: jeevanullas at gmail.com (Deependra Shekhawat) Date: Mon, 5 Mar 2007 10:12:36 +0530 Subject: [et-mgmt-tools] Re: domain creation problem in virt-manager 0.3.1 In-Reply-To: <57127b5f0703042023h7aa20a0dha056372b2a522edf@mail.gmail.com> References: <57127b5f0703042023h7aa20a0dha056372b2a522edf@mail.gmail.com> Message-ID: <57127b5f0703042042t5ec1c6d5ge74937c49226656e@mail.gmail.com> Okay. I figured out the problem. It was python-virtinst conflicting with virtinst-0.101.0. So I un-installed python-virtinst and copied the required files from virtinst-0.101.0 to the /usr/lib/python-2.4/site-packages/virtinst and /usr/sbin/virt-install. Now I can click on the New button and it works. I feeded in the values but when I clicked on finish button it said: Guest Install Error ERROR: [Errno 2] No such file or directory: '/var/lib/xen/virtinstmnt.NAku6T/RPM-GPG-KEY' Please Help. Regards Deependra Shekhawat RHCE On 3/5/07, Deependra Shekhawat wrote: > > Hello, > I am using virt-manager 0.3.1. I tried to create a new domU using the > virt-manager but got the following error: > [root at server ~]# virt-manager > GTK Accessibility Module initialized > Traceback (most recent call last): > File "/usr/share/virt-manager/virtManager/engine.py", line 161, in > _do_show_create > self.show_create(uri) > File "/usr/share/virt-manager/virtManager/engine.py", line 253, in > show_create > self.windowCreate.reset_state() > File "/usr/share/virt-manager/virtManager/create.py", line 269, in > reset_state > os_list = virtinst.FullVirtGuest.OS_TYPES.keys() > AttributeError: type object 'FullVirtGuest' has no attribute 'OS_TYPES' > > Is it a known bug? > I have libvirt-0.2.0-1, libvirt-python-0.2.0-1, virtinst-0.101.0 and > python-virtinst-0.96.0-2.el5. > > Regards > Deependra Shekhawat > RHCE > > -- Enjoy Life ! -------------- next part -------------- An HTML attachment was scrubbed... URL: From jeevanullas at gmail.com Mon Mar 5 04:23:17 2007 From: jeevanullas at gmail.com (Deependra Shekhawat) Date: Mon, 5 Mar 2007 09:53:17 +0530 Subject: [et-mgmt-tools] domain creation problem in virt-manager 0.3.1 Message-ID: <57127b5f0703042023h7aa20a0dha056372b2a522edf@mail.gmail.com> Hello, I am using virt-manager 0.3.1. I tried to create a new domU using the virt-manager but got the following error: [root at server ~]# virt-manager GTK Accessibility Module initialized Traceback (most recent call last): File "/usr/share/virt-manager/virtManager/engine.py", line 161, in _do_show_create self.show_create(uri) File "/usr/share/virt-manager/virtManager/engine.py", line 253, in show_create self.windowCreate.reset_state() File "/usr/share/virt-manager/virtManager/create.py", line 269, in reset_state os_list = virtinst.FullVirtGuest.OS_TYPES.keys() AttributeError: type object 'FullVirtGuest' has no attribute 'OS_TYPES' Is it a known bug? I have libvirt-0.2.0-1, libvirt-python-0.2.0-1, virtinst-0.101.0 and python-virtinst-0.96.0-2.el5. Regards Deependra Shekhawat RHCE -------------- next part -------------- An HTML attachment was scrubbed... URL: From tforeman at ibsys.com Mon Mar 5 15:33:40 2007 From: tforeman at ibsys.com (Foreman, Tim) Date: Mon, 5 Mar 2007 09:33:40 -0600 Subject: [et-mgmt-tools] virt-name in cobbler profile? Message-ID: <514B41B4EB0E834EAFB17FA73E263B7009B2FF@exch001.ibsys.com> The man page for cobbler says: virt-name (optional) (Virt-only) what the Virt guest name should start with. Creating multiple images on a machine will cause increasing numbers to be appended to this name. The default is "virtguest". When I specify a virt-name of RH4U5 and try and create multiple guests using koan: koan --virt --server=kickstart.mtc.ibsys.com \ --profile=rhel4u5-beta_xen_domU It creates the first one named RH4U5 but does not append a number to the second one and fails with an error: libvir: Xen Daemon error : POST operation failed: (xend.err "Error creating domain: VM name 'RH4U5' already in use by domain 2") Is this functionailty missing? Not implemented yet? Also the koan man page says: If you add the --interactive flag, you can watch the kickstart as it happens in an "xm console". But koan says it doesn't know what --interactive means: koan: error: no such option: --interactive -- Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com -- The Onion: Have you decided what you want to be when you grow up? Berkeley Breathed: Dad. The rest is frosting. From mdehaan at redhat.com Mon Mar 5 15:58:33 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 05 Mar 2007 10:58:33 -0500 Subject: [et-mgmt-tools] virt-name in cobbler profile? In-Reply-To: <514B41B4EB0E834EAFB17FA73E263B7009B2FF@exch001.ibsys.com> References: <514B41B4EB0E834EAFB17FA73E263B7009B2FF@exch001.ibsys.com> Message-ID: <45EC3E29.2010004@redhat.com> Foreman, Tim wrote: > The man page for cobbler says: > > virt-name > (optional) (Virt-only) what the Virt guest name should start with. > Creating multiple images on a machine will cause increasing numbers > to be appended to this name. The default is "virtguest". > > When I specify a virt-name of RH4U5 and try and create multiple > guests using koan: > koan --virt --server=kickstart.mtc.ibsys.com \ > --profile=rhel4u5-beta_xen_domU > > It creates the first one named RH4U5 but does not append a number > to the second one and fails with an error: > > libvir: Xen Daemon error : POST operation failed: (xend.err "Error > creating domain: VM name 'RH4U5' already in use by domain 2") > > Is this functionailty missing? Not implemented yet? > This is probably a bug. koan keeps track of Xen names that are already taken by looking at files in /etc/xen, so let me know if you see a /etc/xen/RH4U5 and /etc/xen/RH4U5_1 in there. I'm guessing koan is confused because the first guest is still installing ... It should be noted that the way koan looks for virtual systems is a bit outdated. The /etc/xen/ paths are deprecated and Xen doesn't use them anymore. Fortunately updating koan is on my short list... > Also the koan man page says: > If you add the --interactive flag, you can watch the kickstart > as it happens in an "xm console". > > But koan says it doesn't know what --interactive means: > koan: error: no such option: --interactive > Yes, koan doesn't support this anymore ... I'll remove this from the manpage. Thanks! > -- > Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com > (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com > -- > The Onion: Have you decided what you want to be when you grow up? > Berkeley Breathed: Dad. The rest is frosting. > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From tforeman at ibsys.com Mon Mar 5 20:46:08 2007 From: tforeman at ibsys.com (Foreman, Tim) Date: Mon, 5 Mar 2007 14:46:08 -0600 Subject: [et-mgmt-tools] virt-name in cobbler profile? Message-ID: <514B41B4EB0E834EAFB17FA73E263B7009B30F@exch001.ibsys.com> > -----Original Message----- > From: et-mgmt-tools-bounces at redhat.com > [mailto:et-mgmt-tools-bounces at redhat.com]On Behalf Of Michael DeHaan > Sent: Monday, March 05, 2007 9:59 AM > To: Fedora/Linux Management Tools > Subject: Re: [et-mgmt-tools] virt-name in cobbler profile? > > > Foreman, Tim wrote: > > The man page for cobbler says: > > > > virt-name > > (optional) (Virt-only) what the Virt guest name should start with. > > Creating multiple images on a machine will cause increasing numbers > > to be appended to this name. The default is "virtguest". > > > > When I specify a virt-name of RH4U5 and try and create multiple > > guests using koan: > > koan --virt --server=kickstart.mtc.ibsys.com \ > > --profile=rhel4u5-beta_xen_domU > > > > It creates the first one named RH4U5 but does not append a number > > to the second one and fails with an error: > > > > libvir: Xen Daemon error : POST operation failed: (xend.err "Error > > creating domain: VM name 'RH4U5' already in use by domain 2") > > > > Is this functionailty missing? Not implemented yet? > > > This is probably a bug. koan keeps track of Xen names that > are already > taken by looking at files in /etc/xen, so let me know if you see a > /etc/xen/RH4U5 and /etc/xen/RH4U5_1 in there. I'm guessing koan is > confused because the first guest is still installing ... > > It should be noted that the way koan looks for virtual > systems is a bit > outdated. The /etc/xen/ paths are deprecated and Xen > doesn't use them > anymore. Fortunately updating koan is on my short list... Ah, the virt-name is my fault. My domain creations were failing, so it wasn't creating the config file, but the domain was running in an odd state. Once I sorted that out it started working. --Tim From mdehaan at redhat.com Mon Mar 5 20:52:26 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 05 Mar 2007 15:52:26 -0500 Subject: [et-mgmt-tools] virt-name in cobbler profile? In-Reply-To: <514B41B4EB0E834EAFB17FA73E263B7009B30F@exch001.ibsys.com> References: <514B41B4EB0E834EAFB17FA73E263B7009B30F@exch001.ibsys.com> Message-ID: <45EC830A.6090502@redhat.com> Foreman, Tim wrote: >> -----Original Message----- >> From: et-mgmt-tools-bounces at redhat.com >> [mailto:et-mgmt-tools-bounces at redhat.com]On Behalf Of Michael DeHaan >> Sent: Monday, March 05, 2007 9:59 AM >> To: Fedora/Linux Management Tools >> Subject: Re: [et-mgmt-tools] virt-name in cobbler profile? >> >> >> Foreman, Tim wrote: >> >>> The man page for cobbler says: >>> >>> virt-name >>> (optional) (Virt-only) what the Virt guest name should start with. >>> Creating multiple images on a machine will cause increasing numbers >>> to be appended to this name. The default is "virtguest". >>> >>> When I specify a virt-name of RH4U5 and try and create multiple >>> guests using koan: >>> koan --virt --server=kickstart.mtc.ibsys.com \ >>> --profile=rhel4u5-beta_xen_domU >>> >>> It creates the first one named RH4U5 but does not append a number >>> to the second one and fails with an error: >>> >>> libvir: Xen Daemon error : POST operation failed: (xend.err "Error >>> creating domain: VM name 'RH4U5' already in use by domain 2") >>> >>> Is this functionailty missing? Not implemented yet? >>> >>> >> This is probably a bug. koan keeps track of Xen names that >> are already >> taken by looking at files in /etc/xen, so let me know if you see a >> /etc/xen/RH4U5 and /etc/xen/RH4U5_1 in there. I'm guessing koan is >> confused because the first guest is still installing ... >> >> It should be noted that the way koan looks for virtual >> systems is a bit >> outdated. The /etc/xen/ paths are deprecated and Xen >> doesn't use them >> anymore. Fortunately updating koan is on my short list... >> > > Ah, the virt-name is my fault. My domain creations were failing, so > it wasn't creating the config file, but the domain was running in an > odd state. > > Once I sorted that out it started working. > > --Tim > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > Glad to hear it worked out ... it's still a bit wrong in koan even so and I'll see if I can't improve that. Meanwhile, as for failed installs, this might be a good idea to do some testing on cobbler's kickstart tracking feature (/var/log/cobbler and "cobbler check"). I haven't heard many field reports on them, but that feature is basically designed to figure out if any kickstarts have stalled/hung/failed. --Michael From mdehaan at redhat.com Mon Mar 5 20:59:56 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 05 Mar 2007 15:59:56 -0500 Subject: [et-mgmt-tools] virt-name in cobbler profile? In-Reply-To: <45EC830A.6090502@redhat.com> References: <514B41B4EB0E834EAFB17FA73E263B7009B30F@exch001.ibsys.com> <45EC830A.6090502@redhat.com> Message-ID: <45EC84CC.5060701@redhat.com> >> > Glad to hear it worked out ... it's still a bit wrong in koan even so > and I'll see if I can't improve that. > > Meanwhile, as for failed installs, this might be a good idea to do > some testing on cobbler's kickstart tracking feature (/var/log/cobbler > and "cobbler check"). I haven't heard many field reports on them, > but that feature is basically designed to figure out if any kickstarts > have stalled/hung/failed. I meant "cobbler status" and not "cobbler check". > > --Michael > > From fj0873gn at aa.jp.fujitsu.com Tue Mar 6 03:30:24 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Tue, 6 Mar 2007 12:30:24 +0900 Subject: [et-mgmt-tools] [PATCH] "--file-size" option is ignored Message-ID: <200703061230.DHG04272.0G849HOK@aa.jp.fujitsu.com> Hi When I execute "virt-install --file-size=10", the value of "--file-size" option is ignored and the error message "ERROR: float() argument must be a string or a number" is output after I answer "What would you like to use as the disk (path)?". The attached patch adds to solve this problem. Signed-off-by: Nobuhiro Itou Thanks, Nobuhiro Itou. Index: virt-install (python-virtinst-0.101.0) =================================================================== --- virt-install 2007-03-05 06:03:21.000000000 +0900 +++ virt-install.disk_size 2007-03-05 19:32:50.000000000 +0900 @@ -99,7 +99,10 @@ def get_vcpus(vcpus, guest): def get_disk(disk, size, sparse, guest, hvm): # FIXME: need to handle a list of disks at some point while 1: - disk = prompt_for_input("What would you like to use as the disk (path)?", disk) + msg = "What would you like to use as the disk (path)?" + if not size is None: + msg = "What would you like the size %sGB to use as the disk (path)?" %(size,) + disk = prompt_for_input(msg, disk) while 1: if os.path.exists(disk): break @@ -131,10 +134,15 @@ def get_disks(disk, size, sparse, guest, sys.exit(1) elif type(disk) == list: size = [ None ] * len(disk) + elif type(size) == list: + disk = [ None ] * len(size) if (type(disk) == list): map(lambda d, s: get_disk(d, s, sparse, guest, hvm), disk, size) + elif (type(size) == list): + map(lambda d, s: get_disk(d, s, sparse, guest, hvm), + disk, size) else: get_disk(disk, size, sparse, guest, hvm) From y.yaegashi at jp.fujitsu.com Tue Mar 6 07:51:26 2007 From: y.yaegashi at jp.fujitsu.com (Yoshihiko Yaegashi) Date: Tue, 06 Mar 2007 16:51:26 +0900 Subject: [et-mgmt-tools] About a solution to operate the virt-manager Message-ID: <45ED1D7E.9070805@jp.fujitsu.com> Hi, I cannot operate the virt-manager in Fedora 7 test2. (on screen of attachment file) It freezes at both X-window and VNC. I seek a solution to the issue. Thanks Yaegashi -------------- next part -------------- A non-text attachment was scrubbed... Name: 01.jpg Type: image/jpeg Size: 89050 bytes Desc: not available URL: From fukuta.saori at jp.fujitsu.com Tue Mar 6 09:09:01 2007 From: fukuta.saori at jp.fujitsu.com (Saori Fukuta) Date: Tue, 06 Mar 2007 18:09:01 +0900 Subject: [et-mgmt-tools] [PATCH] check of the required arguments Message-ID: <20070306180056.6F31.FUKUTA.SAORI@jp.fujitsu.com> Hi, I can run virt-install with following option arguments: # virt-install --name # virt-install --name= but virt-install prints out a different message. For example: 1) virt-install with "--name" # virt-install --name Usage: virt-install [options] virt-install: error: --name option requires an argument 2) virt-install with "--name=" but no argument # virt-install --name= Would you like a fully virtualized guest (yes or no)? The cause is parsing for command line options by optparse. If the command-line being parsed includes "--name=", then optparse, on seeing "--name" option, set '' as a value. I know long option allows following arguments: --file=foo --file foo but I think it's better if each operation print same message. So, here's what I'm going to suggest. I add two functions that check the length of values, and if the value type is string I defined them as a callback options . They are called by following option: long option (short option) --name (-n) --uuid (-u) --file (-f) --mac (-m) --bridge (-b) --connect --cdrom (-c) --os-type --os-variant --arch --location (-l) There is one exception. "--extra-args" doesn't call, because the option has default value and that is ''. For example ( after fix ): 1) virt-install with "--name" # virt-install --name Usage: virt-install [options] virt-install: error: --name option requires an argument 2) virt-install with "--name=" but no argument # virt-install --name= Usage: virt-install [options] virt-install: error: --name option requires an argument And, these have done with the patch of changeset 110(Further improvements to name validation) ;-) Signed-off-by: Saori Fukuta Thanks, Saori Fukuta. Index: virt-install =================================================================== diff -r 3e18fa0cafc4 virt-install --- a/virt-install Fri Mar 02 09:16:33 2007 -0500 +++ b/virt-install Tue Mar 06 14:22:20 2007 +0900 @@ -15,7 +15,7 @@ import os, sys, string -from optparse import OptionParser +from optparse import OptionParser, OptionValueError import subprocess import struct import logging @@ -213,20 +213,32 @@ def get_fullvirt_cdrom(cdpath, guest): cdpath = None ### Option parsing +def check_before_store(option, opt_str, value, parser): + if len(value) == 0: + raise OptionValueError, "%s option requires an argument" %opt_str + setattr(parser.values, option.dest, value) + +def check_before_append(option, opt_str, value, parser): + if len(value) == 0: + raise OptionValueError, "%s option requires an argument" %opt_str + parser.values.ensure_value(option.dest, []).append(value) + def parse_args(): parser = OptionParser() parser.add_option("-n", "--name", type="string", dest="name", + action="callback", callback=check_before_store, help="Name of the guest instance") parser.add_option("-r", "--ram", type="int", dest="memory", help="Memory to allocate for guest instance in megabytes") parser.add_option("-u", "--uuid", type="string", dest="uuid", + action="callback", callback=check_before_store, help="UUID for the guest; if none is given a random UUID will be generated") parser.add_option("", "--vcpus", type="int", dest="vcpus", help="Number of vcpus to configure for your guest") # disk options - parser.add_option("-f", "--file", type="string", - dest="diskfile", action="append", + parser.add_option("-f", "--file", type="string", dest="diskfile", + action="callback", callback=check_before_append, help="File to use as the disk image") parser.add_option("-s", "--file-size", type="float", action="append", dest="disksize", @@ -236,11 +248,11 @@ def parse_args(): help="Don't use sparse files for disks. Note that this will be significantly slower for guest creation") # network options - parser.add_option("-m", "--mac", type="string", - dest="mac", action="append", + parser.add_option("-m", "--mac", type="string", dest="mac", + action="callback", callback=check_before_append, help="Fixed MAC address for the guest; if none or RANDOM is given a random address will be used") - parser.add_option("-b", "--bridge", type="string", - dest="bridge", action="append", + parser.add_option("-b", "--bridge", type="string", dest="bridge", + action="callback", callback=check_before_append, help="Bridge to connect guest NIC to; if none given, will try to determine the default") # graphics options @@ -259,23 +271,32 @@ def parse_args(): parser.add_option("", "--accelerate", action="store_true", dest="accelerate", help="Use kernel acceleration capabilities") parser.add_option("", "--connect", type="string", dest="connect", + action="callback", callback=check_before_store, help="Connect to hypervisor with URI") # fullvirt options parser.add_option("-v", "--hvm", action="store_true", dest="fullvirt", help="This guest should be a fully virtualized guest") parser.add_option("-c", "--cdrom", type="string", dest="cdrom", + action="callback", callback=check_before_store, help="File to use a virtual CD-ROM device for fully virtualized guests") - parser.add_option("", "--os-type", type="string", dest="os_type", help="The OS type for fully virtualized guests, e.g. Linux, Solaris, Windows", default="Other") - parser.add_option("", "--os-variant", type="string", dest="os_variant", help="The OS variant for fully virtualized guests, e.g. Fedora, Solaris 8, Windows XP", default="Other") + parser.add_option("", "--os-type", type="string", dest="os_type", + action="callback", callback=check_before_store, + help="The OS type for fully virtualized guests, e.g. Linux, Solaris, Windows", default="Other") + parser.add_option("", "--os-variant", type="string", dest="os_variant", + action="callback", callback=check_before_store, + help="The OS variant for fully virtualized guests, e.g. Fedora, Solaris 8, Windows XP", default="Other") parser.add_option("", "--noapic", action="store_true", dest="noapic", help="Disables APIC for fully virtualized guest (overrides value in os-type/os-variant db)", default=False) parser.add_option("", "--noacpi", action="store_true", dest="noacpi", help="Disables ACPI for fully virtualized guest (overrides value in os-type/os-variant db)", default=False) - parser.add_option("", "--arch", type="string", dest="arch", help="The CPU architecture to simulate") + parser.add_option("", "--arch", type="string", dest="arch", + action="callback", callback=check_before_store, + help="The CPU architecture to simulate") # paravirt options parser.add_option("-p", "--paravirt", action="store_false", dest="fullvirt", help="This guest should be a paravirtualized guest") parser.add_option("-l", "--location", type="string", dest="location", + action="callback", callback=check_before_store, help="Installation source for paravirtualized guest (eg, nfs:host:/path, http://host/path, ftp://host/path)") parser.add_option("-x", "--extra-args", type="string", dest="extra", default="", From berrange at redhat.com Tue Mar 6 12:15:53 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Tue, 6 Mar 2007 12:15:53 +0000 Subject: [et-mgmt-tools] About a solution to operate the virt-manager In-Reply-To: <45ED1D7E.9070805@jp.fujitsu.com> References: <45ED1D7E.9070805@jp.fujitsu.com> Message-ID: <20070306121553.GB21807@redhat.com> On Tue, Mar 06, 2007 at 04:51:26PM +0900, Yoshihiko Yaegashi wrote: > Hi, > > I cannot operate the virt-manager in Fedora 7 test2. > (on screen of attachment file) > It freezes at both X-window and VNC. > > I seek a solution to the issue. This is a GNOME / GTK bug I believe - try disabling 'Accessibility' support in GNOME, and the logging out from X and strting a new GNOME desktop session. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From takatom at jp.fujitsu.com Tue Mar 6 12:59:16 2007 From: takatom at jp.fujitsu.com (Takahashi Tomohiro) Date: Tue, 6 Mar 2007 21:59:16 +0900 Subject: [et-mgmt-tools] [PATCH][RFC] Add --keymap option to virt-install Message-ID: <38e301c75fef$63a35720$ecb1220a@fj98229700> Hi, I would like to add --keymap option to virt-install in order to set up jp106 keymap. Because I can't use jp106 keyboard. So, I made a patch (notes:It relates libvirt) Usage: virt-install --keymap=ja I confirmed that I can use jp106 keyboard on HVM domain. But didn't confirm PV domain. Please give me an advice, if you have it. Signed-off-by: Tomohiro Takahashi Thanks, Tomohiro Takahashi. =================================================================== diff -uNrp virtinst-0.101.0.orig/virt-install virtinst-0.101.0/virt-install --- virtinst-0.101.0.orig/virt-install 2007-03-05 21:29:44.000000000 +0900 +++ virtinst-0.101.0/virt-install 2007-03-06 15:51:22.000000000 +0900 @@ -96,6 +96,13 @@ def get_vcpus(vcpus, guest): except ValueError, e: print "ERROR: ", e +def get_keymap(keymap, guest): + if keymap: + try: + guest.keymap = keymap + except ValueError, e: + print "ERROR: ", e + def get_disk(disk, size, sparse, guest, hvm): # FIXME: need to handle a list of disks at some point while 1: @@ -160,14 +167,14 @@ def get_networks(macs, bridges, guest): else: get_network(macs, bridges, guest) -def get_graphics(vnc, vncport, nographics, sdl, guest): +def get_graphics(vnc, vncport, nographics, sdl, keymap, guest): if vnc and nographics: raise ValueError, "Can't do both VNC graphics and nographics" if nographics: guest.graphics = False return if vnc is not None: - guest.graphics = (True, "vnc", vncport) + guest.graphics = (True, "vnc", vncport, keymap) return if sdl is not None: guest.graphics = (True, "sdl") @@ -180,7 +187,7 @@ def get_graphics(vnc, vncport, nographic print "ERROR", e continue if vnc: - guest.graphics = "vnc" + guest.graphics = (True, "vnc", vncport, keymap) else: guest.graphics = False break @@ -256,6 +263,9 @@ def parse_args(): action="store_false", dest="autoconsole", help="Don't automatically try to connect to the guest console") + parser.add_option("-k", "--keymap", type="string", dest="keymap", + help="set up keymap for a graphical console") + parser.add_option("", "--accelerate", action="store_true", dest="accelerate", help="Use kernel acceleration capabilities") parser.add_option("", "--connect", type="string", dest="connect", @@ -417,6 +427,7 @@ def main(): get_memory(options.memory, guest) get_uuid(options.uuid, guest) get_vcpus(options.vcpus, guest) + get_keymap(options.keymap, guest) # set up disks get_disks(options.diskfile, options.disksize, options.sparse, @@ -426,7 +437,7 @@ def main(): get_networks(options.mac, options.bridge, guest) # set up graphics information - get_graphics(options.vnc, options.vncport, options.nographics, options.sdl, guest) + get_graphics(options.vnc, options.vncport, options.nographics, options.sdl, options.keymap, guest) # and now for the full-virt vs paravirt specific questions if not hvm: # paravirt diff -uNrp virtinst-0.101.0.orig/virtinst/Guest.py virtinst-0.101.0/virtinst/Guest.py --- virtinst-0.101.0.orig/virtinst/Guest.py 2007-03-05 21:29:44.000000000 +0900 +++ virtinst-0.101.0/virtinst/Guest.py 2007-03-06 15:39:08.000000000 +0900 @@ -185,9 +185,17 @@ class VNCVirtualGraphics(XenGraphics): self.port = args[0] else: self.port = -1 + if len(args) >= 2 and args[1]: + self.keymap = args[1] + else: + self.keymap = None def get_xml_config(self): - return " " % (self.port) + if self.keymap == None: + keymapstr = "" + else: + keymapstr = "keymap='"+self.keymap+"' " + return " " % {"port":self.port, "keymapstr":keymapstr} # Back compat class to avoid ABI break class XenVNCGraphics(VNCVirtualGraphics): @@ -218,6 +226,7 @@ class Guest(object): self._maxmemory = None self._vcpus = None self._graphics = { "enabled": False } + self._keymap = None self.domain = None self.conn = connection @@ -301,6 +310,21 @@ class Guest(object): self._vcpus = val vcpus = property(get_vcpus, set_vcpus) + + # keymap for the guest + def get_keymap(self): + return self._keymap + def set_keymap(self, val): + if val and (re.match("^[a-zA-Z0-9_]*$", val) == None): + raise ValueError, "Keymap be alphanumeric or _" + if val and (len(val) > 16): + raise ValueError, "Keymap must be less than 16 characters" + if val and (type(val) != type("string")): + raise ValueError, "Keymap must be a string" + self._keymap = val + keymap = property(get_keymap, set_keymap) + + # kernel + initrd pair to use for installing as opposed to using a location def get_boot(self): return self._boot @@ -374,6 +398,7 @@ class Guest(object): if len(val) >= 1: self._graphics["enabled"] = val[0] if len(val) >= 2: t = val[1] if len(val) >= 3: opts = val[2] + if len(val) >= 4: self._graphics["keymap"] = val[3] else: if val in ("vnc", "sdl"): t = val @@ -386,7 +411,7 @@ class Guest(object): if self._graphics["enabled"] == True: if t == "vnc": - gt = VNCVirtualGraphics(opts) + gt = VNCVirtualGraphics(opts, self._graphics["keymap"]) elif t == "sdl": gt = SDLVirtualGraphics(opts) else: From takatom at tokyo.ssg.fujitsu.com Tue Mar 6 12:49:16 2007 From: takatom at tokyo.ssg.fujitsu.com (Takahashi Tomohiro) Date: Tue, 6 Mar 2007 21:49:16 +0900 Subject: [et-mgmt-tools] [PATCH][RFC] Add --keymap option to virt-install Message-ID: <38c801c75fee$39157070$ecb1220a@fj98229700> Hi, I would like to add --keymap option to virt-install in order to set up jp106 keymap. Because I can't use jp106 keyboard. So, I made a patch (notes:It relates libvirt) Usage: virt-install --keymap=ja I confirmed that I can use jp106 keyboard on HVM domain. But didn't confirm PV domain. Please give me an advice, if you have it. Signed-off-by: Tomohiro Takahashi Thanks, Tomohiro Takahashi. =================================================================== diff -uNrp virtinst-0.101.0.orig/virt-install virtinst-0.101.0/virt-install --- virtinst-0.101.0.orig/virt-install 2007-03-05 21:29:44.000000000 +0900 +++ virtinst-0.101.0/virt-install 2007-03-06 15:51:22.000000000 +0900 @@ -96,6 +96,13 @@ def get_vcpus(vcpus, guest): except ValueError, e: print "ERROR: ", e +def get_keymap(keymap, guest): + if keymap: + try: + guest.keymap = keymap + except ValueError, e: + print "ERROR: ", e + def get_disk(disk, size, sparse, guest, hvm): # FIXME: need to handle a list of disks at some point while 1: @@ -160,14 +167,14 @@ def get_networks(macs, bridges, guest): else: get_network(macs, bridges, guest) -def get_graphics(vnc, vncport, nographics, sdl, guest): +def get_graphics(vnc, vncport, nographics, sdl, keymap, guest): if vnc and nographics: raise ValueError, "Can't do both VNC graphics and nographics" if nographics: guest.graphics = False return if vnc is not None: - guest.graphics = (True, "vnc", vncport) + guest.graphics = (True, "vnc", vncport, keymap) return if sdl is not None: guest.graphics = (True, "sdl") @@ -180,7 +187,7 @@ def get_graphics(vnc, vncport, nographic print "ERROR", e continue if vnc: - guest.graphics = "vnc" + guest.graphics = (True, "vnc", vncport, keymap) else: guest.graphics = False break @@ -256,6 +263,9 @@ def parse_args(): action="store_false", dest="autoconsole", help="Don't automatically try to connect to the guest console") + parser.add_option("-k", "--keymap", type="string", dest="keymap", + help="set up keymap for a graphical console") + parser.add_option("", "--accelerate", action="store_true", dest="accelerate", help="Use kernel acceleration capabilities") parser.add_option("", "--connect", type="string", dest="connect", @@ -417,6 +427,7 @@ def main(): get_memory(options.memory, guest) get_uuid(options.uuid, guest) get_vcpus(options.vcpus, guest) + get_keymap(options.keymap, guest) # set up disks get_disks(options.diskfile, options.disksize, options.sparse, @@ -426,7 +437,7 @@ def main(): get_networks(options.mac, options.bridge, guest) # set up graphics information - get_graphics(options.vnc, options.vncport, options.nographics, options.sdl, guest) + get_graphics(options.vnc, options.vncport, options.nographics, options.sdl, options.keymap, guest) # and now for the full-virt vs paravirt specific questions if not hvm: # paravirt diff -uNrp virtinst-0.101.0.orig/virtinst/Guest.py virtinst-0.101.0/virtinst/Guest.py --- virtinst-0.101.0.orig/virtinst/Guest.py 2007-03-05 21:29:44.000000000 +0900 +++ virtinst-0.101.0/virtinst/Guest.py 2007-03-06 15:39:08.000000000 +0900 @@ -185,9 +185,17 @@ class VNCVirtualGraphics(XenGraphics): self.port = args[0] else: self.port = -1 + if len(args) >= 2 and args[1]: + self.keymap = args[1] + else: + self.keymap = None def get_xml_config(self): - return " " % (self.port) + if self.keymap == None: + keymapstr = "" + else: + keymapstr = "keymap='"+self.keymap+"' " + return " " % {"port":self.port, "keymapstr":keymapstr} # Back compat class to avoid ABI break class XenVNCGraphics(VNCVirtualGraphics): @@ -218,6 +226,7 @@ class Guest(object): self._maxmemory = None self._vcpus = None self._graphics = { "enabled": False } + self._keymap = None self.domain = None self.conn = connection @@ -301,6 +310,21 @@ class Guest(object): self._vcpus = val vcpus = property(get_vcpus, set_vcpus) + + # keymap for the guest + def get_keymap(self): + return self._keymap + def set_keymap(self, val): + if val and (re.match("^[a-zA-Z0-9_]*$", val) == None): + raise ValueError, "Keymap be alphanumeric or _" + if val and (len(val) > 16): + raise ValueError, "Keymap must be less than 16 characters" + if val and (type(val) != type("string")): + raise ValueError, "Keymap must be a string" + self._keymap = val + keymap = property(get_keymap, set_keymap) + + # kernel + initrd pair to use for installing as opposed to using a location def get_boot(self): return self._boot @@ -374,6 +398,7 @@ class Guest(object): if len(val) >= 1: self._graphics["enabled"] = val[0] if len(val) >= 2: t = val[1] if len(val) >= 3: opts = val[2] + if len(val) >= 4: self._graphics["keymap"] = val[3] else: if val in ("vnc", "sdl"): t = val @@ -386,7 +411,7 @@ class Guest(object): if self._graphics["enabled"] == True: if t == "vnc": - gt = VNCVirtualGraphics(opts) + gt = VNCVirtualGraphics(opts, self._graphics["keymap"]) elif t == "sdl": gt = SDLVirtualGraphics(opts) else: From berrange at redhat.com Tue Mar 6 15:28:10 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Tue, 6 Mar 2007 15:28:10 +0000 Subject: [et-mgmt-tools] [PATCH][RFC] Add --keymap option to virt-install In-Reply-To: <38e301c75fef$63a35720$ecb1220a@fj98229700> References: <38e301c75fef$63a35720$ecb1220a@fj98229700> Message-ID: <20070306152810.GE29872@redhat.com> On Tue, Mar 06, 2007 at 09:59:16PM +0900, Takahashi Tomohiro wrote: > Hi, > > I would like to add --keymap option to virt-install in order to set up > jp106 keymap. Because I can't use jp106 keyboard. So, I made a patch > (notes:It relates libvirt) > > Usage: virt-install --keymap=ja > > I confirmed that I can use jp106 keyboard on HVM domain. > But didn't confirm PV domain. > > Please give me an advice, if you have it. ACK. The patch looks fine to me. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From tforeman at ibsys.com Tue Mar 6 16:00:55 2007 From: tforeman at ibsys.com (Foreman, Tim) Date: Tue, 6 Mar 2007 10:00:55 -0600 Subject: [et-mgmt-tools] More Cobbler, Koan and Xen questions Message-ID: <514B41B4EB0E834EAFB17FA73E263B7009B322@exch001.ibsys.com> I'm still trying to kickstart a Xen guest using Cobbler and Koan and I'm making progress, but it's still not working right. I have a distro, I have a profile and I have a system. The problem is that it appears to be ignoring my kickstart file and goes into an interactive install. I see in the webserver logs on the Cobbler box that it's requesting, and getting, the kickstart file, but it appears to not be using it. My question is this: The cobbler man page states that I should use the MAC address for the system name when kickstarting a virtual server, but the MAC address is randomly generated when I run Koan to do a virtual install. How is this supposed to work? -- Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com -- The Onion: Have you decided what you want to be when you grow up? Berkeley Breathed: Dad. The rest is frosting. From mdehaan at redhat.com Tue Mar 6 17:03:18 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 06 Mar 2007 12:03:18 -0500 Subject: [et-mgmt-tools] More Cobbler, Koan and Xen questions In-Reply-To: <514B41B4EB0E834EAFB17FA73E263B7009B322@exch001.ibsys.com> References: <514B41B4EB0E834EAFB17FA73E263B7009B322@exch001.ibsys.com> Message-ID: <45ED9ED6.30809@redhat.com> Foreman, Tim wrote: > I'm still trying to kickstart a Xen guest using Cobbler and Koan > and I'm making progress, but it's still not working right. > > I have a distro, I have a profile and I have a system. The problem > is that it appears to be ignoring my kickstart file and goes into > an interactive install. I see in the webserver logs on the Cobbler > box that it's requesting, and getting, the kickstart file, but it > appears to not be using it. > The really best way to get what you want is to do a manual install using either virtguest-install or virt-manager, and then look at the /root/anaconda.ks and use that as a basis, just like you would for any other system. When kickstarts go interactive, it's because they miss an option they need. A kickstart file is basically an answer file for most of the install process. > My question is this: The cobbler man page states that I should use > the MAC address for the system name when kickstarting a virtual > server, but the MAC address is randomly generated when I run Koan > to do a virtual install. > Well, you don't technically have to, and I should probably clear that up in the manpage. The reason you would want to use the MAC address is so that you could do per-system templating, control the MAC address rather than getting a random one, and the like. This is how you tell koan to provision virt based on a system definition: koan --virt --system=AA:BB:CC:DD:EE:FF --server=foo.example.com This assumes you've done a "cobbler system add --name=AA:BB:CC:DD:EE:FF --profile=virtualfoo" before. > -- > Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com > (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com > -- > The Onion: Have you decided what you want to be when you grow up? > Berkeley Breathed: Dad. The rest is frosting. > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From tforeman at ibsys.com Tue Mar 6 17:31:32 2007 From: tforeman at ibsys.com (Foreman, Tim) Date: Tue, 6 Mar 2007 11:31:32 -0600 Subject: [et-mgmt-tools] More Cobbler, Koan and Xen questions Message-ID: <514B41B4EB0E834EAFB17FA73E263B7009B328@exch001.ibsys.com> > -----Original Message----- > From: et-mgmt-tools-bounces at redhat.com > [mailto:et-mgmt-tools-bounces at redhat.com]On Behalf Of Michael DeHaan > Sent: Tuesday, March 06, 2007 11:03 AM > To: Fedora/Linux Management Tools > Subject: Re: [et-mgmt-tools] More Cobbler, Koan and Xen questions > > Foreman, Tim wrote: > > I'm still trying to kickstart a Xen guest using Cobbler and Koan > > and I'm making progress, but it's still not working right. > > > > I have a distro, I have a profile and I have a system. The problem > > is that it appears to be ignoring my kickstart file and goes into > > an interactive install. I see in the webserver logs on the Cobbler > > box that it's requesting, and getting, the kickstart file, but it > > appears to not be using it. > > > The really best way to get what you want is to do a manual > install using either virtguest-install or virt-manager, > and then look at the /root/anaconda.ks and use that as a basis, just > like you would for any other system. Right. The first time that it started up in interactive mode I did run through it. There were no options in the anaconda.ks file that I don't have in my kickstart file. My kickstart file is a kind of complicated one with some scripting and templating, but it's essentially a tweaked copy of the file that I have been literally using for years to kickstart servers. Both from CD installs and from Cobbler installs using PXE boot. The only things that are different are the fact that I'm using Xen and that I'm trying to start the process with Koan instead of a PXE boot. > When kickstarts go interactive, it's because they miss an option they > need. A kickstart file is basically an answer file > for most of the install process. > > My question is this: The cobbler man page states that I should use > > the MAC address for the system name when kickstarting a virtual > > server, but the MAC address is randomly generated when I run Koan > > to do a virtual install. > > > Well, you don't technically have to, and I should probably > clear that up in the manpage. The reason you would want to use the MAC > address is so that you could do per-system templating, control the MAC > address rather than getting a random one, and the like. > > This is how you tell koan to provision virt based on a system > definition: > > koan --virt --system=AA:BB:CC:DD:EE:FF --server=foo.example.com > > This assumes you've done a "cobbler system add > --name=AA:BB:CC:DD:EE:FF > --profile=virtualfoo" before. I did a "cobbler system add --name=10.9.90.11 --profile=virtualfoo" and used the IP address I wanted the system to end up with. The cobbler man page states you can use a MAC, an IP address or a resolvable hostname for a system name. Then I did a "koan --virt --server=servername --system=10.9.90.11" Maybe what I need to do is run a manual install with no kickstart file and see what's in that anaconda.ks. I might be missing something still. -- Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com -- The Onion: Have you decided what you want to be when you grow up? Berkeley Breathed: Dad. The rest is frosting. From mdehaan at redhat.com Tue Mar 6 18:04:22 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 06 Mar 2007 13:04:22 -0500 Subject: [et-mgmt-tools] More Cobbler, Koan and Xen questions In-Reply-To: <514B41B4EB0E834EAFB17FA73E263B7009B328@exch001.ibsys.com> References: <514B41B4EB0E834EAFB17FA73E263B7009B328@exch001.ibsys.com> Message-ID: <45EDAD26.4050606@redhat.com> Yup, that's true. > Then I did a "koan --virt --server=servername --system=10.9.90.11" > Sure, IP's are fine too. Even though koan has nothing to do with PXE, the requirement for "--system" parameters to be resolveable to IP's or MAC's is grounded in that. So one or the other and you are good to go. > Maybe what I need to do is run a manual install with no kickstart file > and see what's in that anaconda.ks. I might be missing something still. > I'm pretty sure that's it. When you do find out, if you can, share what you found -- somebody else might have the same problem. From tforeman at ibsys.com Tue Mar 6 19:20:54 2007 From: tforeman at ibsys.com (Foreman, Tim) Date: Tue, 6 Mar 2007 13:20:54 -0600 Subject: [et-mgmt-tools] More Cobbler, Koan and Xen questions Message-ID: <514B41B4EB0E834EAFB17FA73E263B7009B32A@exch001.ibsys.com> > -----Original Message----- > From: et-mgmt-tools-bounces at redhat.com > [mailto:et-mgmt-tools-bounces at redhat.com]On Behalf Of Michael DeHaan > Sent: Tuesday, March 06, 2007 12:04 PM > To: Fedora/Linux Management Tools > Subject: Re: [et-mgmt-tools] More Cobbler, Koan and Xen questions > > > Yup, that's true. >> Then I did a "koan --virt --server=servername --system=10.9.90.11" >> > Sure, IP's are fine too. > Even though koan has nothing to do with PXE, the requirement for > "--system" parameters to be resolveable to IP's or MAC's is > grounded in that. So one or the other and you are good to go. > >> Maybe what I need to do is run a manual install with no >> kickstart file and see what's in that anaconda.ks. I might be >> missing something still. > > I'm pretty sure that's it. When you do find out, if you can, share > what you found -- somebody else might have the same problem. Well, that's interesting. Explain this if you can. :-) I created a new profile that uses the empty default.ks file. I added a system that uses that profile. I ran koan and did an interactive install. Then I copied the ananconda-ks.cfg file from that sucessful install to the cobbler box. I edited the anaconda-ks.cfg file to change the hostname and IP address and to enable the disk partitioning section. That's all I changed. Then I created a new profile to use this anaconda-ks.cfg file and added a new system using that profile. I kicked off another koan session using this new profile and system. Guess what it did? It started running an interactive install again! I see in the web server log that it is requesting the ks.cfg file, and I looked at the file it's requesting and saw that it has the right info in it. I also can't figure out how to make 'cobbler status' work. I have a cobbler_track dir, I path everything through it, and 'cobbler status' returns no data. Any ideas? -- Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com -- The Onion: Have you decided what you want to be when you grow up? Berkeley Breathed: Dad. The rest is frosting. From dmourati at gmail.com Tue Mar 6 19:47:02 2007 From: dmourati at gmail.com (Demetri Mouratis) Date: Tue, 6 Mar 2007 11:47:02 -0800 Subject: [et-mgmt-tools] Using cobbler to mirror repos like updates/extras -- further information In-Reply-To: <45DF3D12.4040006@redhat.com> References: <45DF1A5F.10501@redhat.com> <20070223183317.GB3332@sliderule.msquared.com.au> <45DF3D12.4040006@redhat.com> Message-ID: <3cc2cbc40703061147u2dc922d6h8f9ac638044cfa98@mail.gmail.com> On 2/23/07, Michael DeHaan wrote: > > What we have here is a failure to communicate :) > > Two kinds of mirrors, basically. Different concepts, used for different > things. > > "cobbler import" puts files into /var/www/cobbler/ks_mirror -- this is a > kickstart tree mirror. These don't need to be updated, and are essential > for doing full automated installations. However since Anaconda /does/ > allow network installs, you don't /have/ to mirror them on your cobbler > server if you already have a good kickstart tree available over http on > your network. In this case, you'd just use "cobbler distro add", and > save yourself the import steps. However, most home users won't have an > fast kickstart tree available to them, and this is why cobbler import > helps you make one. > > cobbler repo add ... puts files into /var/www/cobbler/repo_mirror -- > these are yum repositories for things like extras & updates. These > update quite frequently and are entirely optional for cobbler -- if you > don't use them, it will use external repos as configured by default in yum. > Hi Michael, I have two requirements, one is a kickstart tree mirror for my base OS, CentOS 4.4. The second requirement is for a repo mirror of a yum repository hosting my companies software setup to be fetched and installed by yum. For the sake of simplicity, I'd like to use repo mirroring for both. As you say above, and I had a surprisingly hard time seeing this, the kickstart tree does not strictly need to be updated. However, is there any harm in doing so? I haven't gone the extra step of changing the /etc/yum.repos.d/* configurations on my target hosts to use my kickstart server as the source for updates but I am considering it. In this case, does it make sense to just start with repo mirroring, treating a repo mirror as a superset of kickstart tree mirroring, and have both the OS and in-house code treated the same in cobbler? Thanks. From mdehaan at redhat.com Tue Mar 6 19:53:56 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 06 Mar 2007 14:53:56 -0500 Subject: [et-mgmt-tools] More Cobbler, Koan and Xen questions In-Reply-To: <514B41B4EB0E834EAFB17FA73E263B7009B32A@exch001.ibsys.com> References: <514B41B4EB0E834EAFB17FA73E263B7009B32A@exch001.ibsys.com> Message-ID: <45EDC6D4.9010603@redhat.com> Foreman, Tim wrote: >> -----Original Message----- >> From: et-mgmt-tools-bounces at redhat.com >> [mailto:et-mgmt-tools-bounces at redhat.com]On Behalf Of Michael DeHaan >> Sent: Tuesday, March 06, 2007 12:04 PM >> To: Fedora/Linux Management Tools >> Subject: Re: [et-mgmt-tools] More Cobbler, Koan and Xen questions >> >> >> Yup, that's true. >> >>> Then I did a "koan --virt --server=servername --system=10.9.90.11" >>> >>> >> Sure, IP's are fine too. >> Even though koan has nothing to do with PXE, the requirement for >> "--system" parameters to be resolveable to IP's or MAC's is >> grounded in that. So one or the other and you are good to go. >> >> >>> Maybe what I need to do is run a manual install with no >>> kickstart file and see what's in that anaconda.ks. I might be >>> missing something still. >>> >> >> I'm pretty sure that's it. When you do find out, if you can, share >> what you found -- somebody else might have the same problem. >> > > Well, that's interesting. Explain this if you can. :-) > > I created a new profile that uses the empty default.ks file. I added a > system that uses that profile. > > I ran koan and did an interactive install. Then I copied the > ananconda-ks.cfg file from that sucessful install to the cobbler box. > > I edited the anaconda-ks.cfg file to change the hostname and IP address > and to enable the disk partitioning section. That's all I changed. > > Then I created a new profile to use this anaconda-ks.cfg file and added > a new system using that profile. > > I kicked off another koan session using this new profile and system. > Guess what it did? > > It started running an interactive install again! > You have to run "cobbler sync" after hand-editing any files, whether those are kickstart files or files in /var/lib/cobbler. So it's still using the old file. > I see in the web server log that it is requesting the ks.cfg file, and > I looked at the file it's requesting and saw that it has the right info > in it. > > I also can't figure out how to make 'cobbler status' work. I have a > cobbler_track dir, I path everything through it, and 'cobbler status' > returns no data. > The "cobbler status" / logging stuff works based on who things: (A) remote syslog data, which currently is set to use port 25150 in /var/lib/cobbler/settings. If a firewall is in the way, that's probably it. (B) file requests to URL's that start with /cobbler_track; these are logged by a mod_python filter handler that cobbler installs. Part (B) is disabled on non-python2.4+ systems because the mod_python filter handler doesn't work very well there. In later versions and Fedora, as well as the coming RHEL5, they should be fine. If for some reason you don't get syslog information, it's likely that "cobbler status" won't return anything because it never saw the install start. If you want to send me the output of "tree /var/log/cobbler" that might be helpful to see if that's the case. > > > -- > Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com > (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com > -- > The Onion: Have you decided what you want to be when you grow up? > Berkeley Breathed: Dad. The rest is frosting. > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From mdehaan at redhat.com Tue Mar 6 20:05:19 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 06 Mar 2007 15:05:19 -0500 Subject: [et-mgmt-tools] Using cobbler to mirror repos like updates/extras -- further information In-Reply-To: <3cc2cbc40703061147u2dc922d6h8f9ac638044cfa98@mail.gmail.com> References: <45DF1A5F.10501@redhat.com> <20070223183317.GB3332@sliderule.msquared.com.au> <45DF3D12.4040006@redhat.com> <3cc2cbc40703061147u2dc922d6h8f9ac638044cfa98@mail.gmail.com> Message-ID: <45EDC97F.4090806@redhat.com> Demetri Mouratis wrote: > On 2/23/07, Michael DeHaan wrote: >> >> What we have here is a failure to communicate :) >> >> Two kinds of mirrors, basically. Different concepts, used for different >> things. >> >> "cobbler import" puts files into /var/www/cobbler/ks_mirror -- this is a >> kickstart tree mirror. These don't need to be updated, and are essential >> for doing full automated installations. However since Anaconda /does/ >> allow network installs, you don't /have/ to mirror them on your cobbler >> server if you already have a good kickstart tree available over http on >> your network. In this case, you'd just use "cobbler distro add", and >> save yourself the import steps. However, most home users won't have an >> fast kickstart tree available to them, and this is why cobbler import >> helps you make one. >> >> cobbler repo add ... puts files into /var/www/cobbler/repo_mirror -- >> these are yum repositories for things like extras & updates. These >> update quite frequently and are entirely optional for cobbler -- if you >> don't use them, it will use external repos as configured by default >> in yum. >> > > > Hi Michael, > > I have two requirements, one is a kickstart tree mirror for my base > OS, CentOS 4.4. The second requirement is for a repo mirror of a yum > repository hosting my companies software setup to be fetched and > installed by yum. For the sake of simplicity, I'd like to use repo > mirroring for both. Sure... > > As you say above, and I had a surprisingly hard time seeing this, the > kickstart tree does not strictly need to be updated. However, is > there any harm in doing so? Well what is imported with "cobbler import --mirror" is an install tree. That tree usually has subdirectories called "os" and "tree". Example: http://download.fedora.redhat.com/pub/fedora/linux/core/6/i386/ Now, what you would mirror in terms of a "core" repo would be a subset of this ... for example, looking at a FC6 system, my fedora-core.repo file references the following URL: http://download.fedora.redhat.com/pub/fedora/linux/core/6/i386/os/ So yes, the repo that gets installed, is something you would have already mirrored with "cobbler import". However, this tree won't change. New packages will appear in the updates repo, and new things may be available in extras or a 3rd party repo, but the repo is the way it was when it shipped without any changes. > I haven't gone the extra step of changing > the /etc/yum.repos.d/* configurations on my target hosts to use my > kickstart server as the source for updates but I am considering it. > In this case, does it make sense to just start with repo mirroring, > treating a repo mirror as a superset of kickstart tree mirroring, and > have both the OS and in-house code treated the same in cobbler? If I were doing this, I would basically let cobbler manage your repos for updates and your companies repo using "cobbler repo add" and "reposync". Then, also in post, I would configure the "core.repo" file to point to http://yourserver/cobbler/ks_mirror/yourdistro/youarch/os" just like the example above indicates. When you run "cobbler reposync" only the updates repo and your company repo would be updated to the cobbler server, but all systems provisioned from it (including the core repo) could still use the cobbler server as an install mirror. > > Thanks. Hope that helped ... if not, let me know. This seems to indicate that it would be useful for cobbler, on systems that are being installed from something in /ks_mirror/ to autoconfigure the "core" repo to use the mirror. Or at least, this should be switchable. Since cobbler doesn't do this automagically now, you can do it in post yourself in the interim. Cobbler knows how to configure the rest of the repos that are managed with "repo add" but not for the original import tree. Eventual consolidation of the repo commands with import seems logical...but "import" is a bit wider grained because it can slurp in distros for multiple arches and so forth. Effectively, cobbler import can slurp in multiple "core" repos at the same time. Since those repos never change, I am hesistant to put them under the domain of "cobbler repo add" and "cobbler reposync". They're really trees, which contain more than the repo. But yeah, autoconfiguration on the client to use the cobbler mirror would be nice. > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From tforeman at ibsys.com Tue Mar 6 20:15:09 2007 From: tforeman at ibsys.com (Foreman, Tim) Date: Tue, 6 Mar 2007 14:15:09 -0600 Subject: [et-mgmt-tools] More Cobbler, Koan and Xen questions Message-ID: <514B41B4EB0E834EAFB17FA73E263B7009B32B@exch001.ibsys.com> > You have to run "cobbler sync" after hand-editing any files, whether > those are kickstart files or files in /var/lib/cobbler. So > it's still using the old file. Of course I did a "cobbler sync". But just to make sure, I ran it again just now and re-ran koan. Well hell, now it's working. I'm 99% sure that I ran "cobbler sync" last time, but now it works. I'll have to test it some more. >> I also can't figure out how to make 'cobbler status' work. I have a >> cobbler_track dir, I path everything through it, and >> 'cobbler status' returns no data. > > The "cobbler status" / logging stuff works based on who things: > > (A) remote syslog data, which currently is set to use port 25150 in > /var/lib/cobbler/settings. If a firewall is in the way, > that's probably it. No firewalls in the way here. > (B) file requests to URL's that start with /cobbler_track; these are > logged by a mod_python filter handler that cobbler installs. > > Part (B) is disabled on non-python2.4+ systems because the mod_python > filter handler doesn't work very well there. In later versions and > Fedora, as well as the coming RHEL5, they should be fine. Ah. I'm on a python 2.3 box, so "cobbler status" isn't going to work for me then. Thanks for the help. -- Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com -- The Onion: Have you decided what you want to be when you grow up? Berkeley Breathed: Dad. The rest is frosting. From dmourati at gmail.com Tue Mar 6 20:32:36 2007 From: dmourati at gmail.com (Demetri Mouratis) Date: Tue, 6 Mar 2007 12:32:36 -0800 Subject: [et-mgmt-tools] Using cobbler to mirror repos like updates/extras -- further information In-Reply-To: <45EDC97F.4090806@redhat.com> References: <45DF1A5F.10501@redhat.com> <20070223183317.GB3332@sliderule.msquared.com.au> <45DF3D12.4040006@redhat.com> <3cc2cbc40703061147u2dc922d6h8f9ac638044cfa98@mail.gmail.com> <45EDC97F.4090806@redhat.com> Message-ID: <3cc2cbc40703061232o59978a02u35f157cad9e1469f@mail.gmail.com> On 3/6/07, Michael DeHaan wrote: > When you run "cobbler reposync" only the updates repo and your company > repo would be updated to the cobbler server, but all systems provisioned > from it (including the core repo) could still use the cobbler server as > an install mirror. > > > > > Thanks. > Hope that helped ... if not, let me know. Hi Michael, Thanks. It makes perfect sense to me. I'm trying to make sure it makes sense to my customers. The finer points about repositories were not something I had given much thought to prior to today's excercise. To be safe, I think I will widen up the scope of what I am syncing and fall back to the mirror command for this purpose. In reality, it seems as though the repository mirroring solution you have put together really shines with the latest Anaconda features as configured at install time. Given that I'm still on 4.4, I believe I may have jumped the gun a bit with repository mirroring. We're managing a bit of this ourselves these days in terms of hacking up entries in /etc/yum.repos.d. Internally, I'd like to see less hacking and more of a finished process around: a, initial installation of the OS b, provisioning the correct repository or repositories for additonal software (we call this bootstrap) c, deployment of addtional software from either initial provisioning server, alternate third party mirror, or internal software mirrors as required. Several folks here have raised concerns about the control issue when production systems are pointing to third party repositories. Doing so potentially removes our ability to stage changes onto our systems in a controlled fashion. This fact leads me to believe that while currently I don't need internal mirrors for all OS/third party software, it may be a good idea to build them in. Thanks for the quick clarification. I think I'm back on track now. -D From tforeman at ibsys.com Tue Mar 6 21:08:32 2007 From: tforeman at ibsys.com (Foreman, Tim) Date: Tue, 6 Mar 2007 15:08:32 -0600 Subject: [et-mgmt-tools] More Cobbler, Koan and Xen questions Message-ID: <514B41B4EB0E834EAFB17FA73E263B7009B32D@exch001.ibsys.com> > -----Original Message----- > From: et-mgmt-tools-bounces at redhat.com > [mailto:et-mgmt-tools-bounces at redhat.com]On Behalf Of Foreman, Tim > Sent: Tuesday, March 06, 2007 2:15 PM > To: 'Fedora/Linux Management Tools' > Subject: RE: [et-mgmt-tools] More Cobbler, Koan and Xen questions >> >> You have to run "cobbler sync" after hand-editing any >> files, whether those are kickstart files or files in >> /var/lib/cobbler. So it's still using the old file. > > Of course I did a "cobbler sync". But just to make sure, I > ran it again just now and re-ran koan. > > Well hell, now it's working. I'm 99% sure that I ran "cobbler > sync" last time, but now it works. > > > > I'll have to test it some more. I take it back. I am seeing totally inconsistant results. I have had it sucessfully perform an automated kickstart exactly once. Deleting the Xen config file and disk image and changing nothing else, and using exactly the same koan command line, I can not get it to perform an automated install. Every time I run it, it starts prompting me for input. Something here is broken. I wonder if I am running into a timing issue where the vif is not up yet, so it can't find the kickstart file... Any other thoughts? Is there anyway to add more debugging output to this process? Is anyone sucessfully using Koan and Cobbler to kickstart Xen guests? -- Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com -- The Onion: Have you decided what you want to be when you grow up? Berkeley Breathed: Dad. The rest is frosting. From mdehaan at redhat.com Tue Mar 6 21:24:30 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 06 Mar 2007 16:24:30 -0500 Subject: [et-mgmt-tools] More Cobbler, Koan and Xen questions In-Reply-To: <514B41B4EB0E834EAFB17FA73E263B7009B32B@exch001.ibsys.com> References: <514B41B4EB0E834EAFB17FA73E263B7009B32B@exch001.ibsys.com> Message-ID: <45EDDC0E.4030205@redhat.com> Foreman, Tim wrote: > > Ah. I'm on a python 2.3 box, so "cobbler status" isn't going to work for > me then. > > Well, technically it should report on the files it logged from syslog anyway, even without the other files being tracked, which is the part that's apparently getting intercepted on your box as well. As I did most of the syslog testing on a FC5 server, that might have something to do with it. The syslog daemon is cobbler_syslogd. > Thanks for the help. > > -- > Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com > (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com > -- > The Onion: Have you decided what you want to be when you grow up? > Berkeley Breathed: Dad. The rest is frosting. > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From mdehaan at redhat.com Tue Mar 6 21:34:07 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 06 Mar 2007 16:34:07 -0500 Subject: [et-mgmt-tools] More Cobbler, Koan and Xen questions In-Reply-To: <514B41B4EB0E834EAFB17FA73E263B7009B32D@exch001.ibsys.com> References: <514B41B4EB0E834EAFB17FA73E263B7009B32D@exch001.ibsys.com> Message-ID: <45EDDE4F.1080509@redhat.com> > > > Is anyone sucessfully using Koan and Cobbler to kickstart Xen > guests? > I haven't tried RHEL4U5 myself ... John Sanabria was using koan for virt a while back. I'm not sure what platform. It might be interesting to see what you get for "virsh list" and "virt dumpxml N" where N is 0 up to the number of virtual machines you have installed. (You may have to install virsh). Also I would be interested in seeing the tree of /var/lib/koan. It's hard to tell from the above description, but it could be that you have a lot of paused/crashed domains that are causing problems for whatever reason. I certaintly don't have any reason as to why some machines would go interactive when others don't, especially during install -- I wouldn't think ram consumption would play a role there. I'm about to put some heavy testing on koan's virt management support for something I'm working on at the moment, so hopefully I can provide some more insight there as well. My apologies on not being more helpful with remote debugging! > -- > Timothy W. Foreman ~ Security Administrator ~ tforeman at ibsys.com > (651) 365-4181 ~ Internet Broadcasting ~ www.ibsys.com > -- > The Onion: Have you decided what you want to be when you grow up? > Berkeley Breathed: Dad. The rest is frosting. > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From mdehaan at redhat.com Tue Mar 6 21:44:28 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 06 Mar 2007 16:44:28 -0500 Subject: [et-mgmt-tools] Using cobbler to mirror repos like updates/extras -- further information In-Reply-To: <3cc2cbc40703061232o59978a02u35f157cad9e1469f@mail.gmail.com> References: <45DF1A5F.10501@redhat.com> <20070223183317.GB3332@sliderule.msquared.com.au> <45DF3D12.4040006@redhat.com> <3cc2cbc40703061147u2dc922d6h8f9ac638044cfa98@mail.gmail.com> <45EDC97F.4090806@redhat.com> <3cc2cbc40703061232o59978a02u35f157cad9e1469f@mail.gmail.com> Message-ID: <45EDE0BC.8020207@redhat.com> Demetri Mouratis wrote: > On 3/6/07, Michael DeHaan wrote: > >> When you run "cobbler reposync" only the updates repo and your company >> repo would be updated to the cobbler server, but all systems provisioned >> from it (including the core repo) could still use the cobbler server as >> an install mirror. >> >> > >> > Thanks. >> Hope that helped ... if not, let me know. > Hi Michael, > > Thanks. > > It makes perfect sense to me. I'm trying to make sure it makes sense > to my customers. The finer points about repositories were not > something I had given much thought to prior to today's excercise. To > be safe, I think I will widen up the scope of what I am syncing and > fall back to the mirror command for this purpose. In reality, it > seems as though the repository mirroring solution you have put > together really shines with the latest Anaconda features as configured > at install time. Given that I'm still on 4.4, I believe I may have > jumped the gun a bit with repository mirroring. If I were you, I would still use the repo mirroring commands. As long as your kickstart does not contain "$yum_repo_stanza", Anaconda will not try to use the repositories, but you are still free to add "$yum_config_stanza" (which is only valid in FC6+/RHEL5+) to automagically configure the contents of /etc/yum.repos.d. Using the repo mirror commands will keep cobbler from being confused, because "cobbler import --mirror" hasn't been designed to import things like "updates". "repo add" has, for instance, it knows to run createrepo to rebuild the repository indexes for files that have been excluded by /etc/cobbler/rsync.exclude. That's pretty important in order to keep your repos in good shape. Once you start installing RHEL5, you can add the "$yum_repo_stanza" bit back in. To see an example of this, diff "/etc/cobbler/kickstart_fc5.ks" with "/etc/cobbler/kickstart_fc6.ks". Not too many differences, just that Anaconda will install packages from updates and make 3rd party repos available in the fc6 one, and in the others, you'll have to do things in post or after the system reboots to get that behavior that ordinarily Anaconda would do for you. > > We're managing a bit of this ourselves these days in terms of hacking > up entries in /etc/yum.repos.d. Internally, I'd like to see less > hacking and more of a finished process around: > > a, initial installation of the OS > b, provisioning the correct repository or repositories for additonal > software (we call this bootstrap) > c, deployment of addtional software from either initial provisioning > server, alternate third party mirror, or internal software mirrors as > required. > > Several folks here have raised concerns about the control issue when > production systems are pointing to third party repositories. Doing so > potentially removes our ability to stage changes onto our systems in a > controlled fashion. This fact leads me to believe that while currently > I don't need internal mirrors for all OS/third party software, it may > be a good idea to build them in. Well, if anything, it makes the installs faster. Probably the right thing to do is to sign your packages and leave the GPGcheck stuff enabled for the repo, but yes, that's true on the staging front. > > Thanks for the quick clarification. I think I'm back on track now. Good deal, hopefully I have not thrown you further off track :) > > -D > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From mdehaan at redhat.com Tue Mar 6 21:53:40 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 06 Mar 2007 16:53:40 -0500 Subject: [et-mgmt-tools] More Cobbler, Koan and Xen questions In-Reply-To: <45EDDE4F.1080509@redhat.com> References: <514B41B4EB0E834EAFB17FA73E263B7009B32D@exch001.ibsys.com> <45EDDE4F.1080509@redhat.com> Message-ID: <45EDE2E4.8010609@redhat.com> > > I haven't tried RHEL4U5 myself ... John Sanabria was using koan for > virt a while back. I'm not sure what platform. > > It might be interesting to see what you get for "virsh list" and "virt > dumpxml N" where N is 0 up to the number of virtual machines you have > installed. (You may have to install virsh). Also I would be > interested in seeing the tree of /var/lib/koan. Apparently it's coded as /var/koan. Told you wrong there. > It's hard to tell from the above description, but it could be that you > have a lot of paused/crashed domains that are causing problems for > whatever reason. I certaintly don't have any reason as to why some > machines would go interactive when others don't, especially during > install -- I wouldn't think ram consumption would play a role there. > > I'm about to put some heavy testing on koan's virt management support > for something I'm working on at the moment, so hopefully I can provide > some more insight there as well. My apologies on not being more > helpful with remote debugging! From dmourati at gmail.com Tue Mar 6 23:09:06 2007 From: dmourati at gmail.com (Demetri Mouratis) Date: Tue, 6 Mar 2007 15:09:06 -0800 Subject: [et-mgmt-tools] Using cobbler to mirror repos like updates/extras -- further information In-Reply-To: <45EDE0BC.8020207@redhat.com> References: <45DF1A5F.10501@redhat.com> <20070223183317.GB3332@sliderule.msquared.com.au> <45DF3D12.4040006@redhat.com> <3cc2cbc40703061147u2dc922d6h8f9ac638044cfa98@mail.gmail.com> <45EDC97F.4090806@redhat.com> <3cc2cbc40703061232o59978a02u35f157cad9e1469f@mail.gmail.com> <45EDE0BC.8020207@redhat.com> Message-ID: <3cc2cbc40703061509j5051f6dgd15afa80eba22c51@mail.gmail.com> Hi Michael, I went with the repo mirroring route: [root at hemops2 ~]# cobbler import --name=centos-4.4 --mirror=rsync://rsync.mirrorservice.org/mirror.centos.org/4.4/os/i386/CentOS/ This went fine. Adding the distro, however, gives me some trouble: [root at hemops2 ~]# cobbler distro add --name=centos4.4 --kernel=/var/www/cobbler/repo_mirror/centos-4.4-i386/images/pxeboot/vmlinuz --initrd=/var/www/cobbler/repo_mirror/centos-4.4-i386/images/pxeboot/initrd.img Traceback (most recent call last): File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 505, in main BootCLI(sys.argv).run() File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 102, in run self.curry_args(self.args[1:], self.commands['toplevel']) File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 421, in curry_args commands[args[0]](args[1:]) File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 469, in distro return self.curry_args(args, self.commands['distro']) File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 421, in curry_args commands[args[0]](args[1:]) File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 386, in distro_edit return self.apply_args(args,commands,on_ok) File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 409, in apply_args on_ok() File "/usr/lib/python2.3/site-packages/cobbler/cobbler.py", line 385, in on_ok = lambda: self.api.distros().add(distro, with_copy=self.api.sync_flag) File "/usr/lib/python2.3/site-packages/cobbler/collection.py", line 97, in add lite_sync.add_single_distro(ref.name) File "/usr/lib/python2.3/site-packages/cobbler/action_litesync.py", line 60, in add_single_distro self.sync.write_distro_file(distro) File "/usr/lib/python2.3/site-packages/cobbler/action_sync.py", line 733, in write_distro_file clone.kernel_options = self.hash_to_string(clone.kernel_options) File "/usr/lib/python2.3/site-packages/cobbler/action_sync.py", line 885, in hash_to_string buffer = buffer + key + "=" + value + " " TypeError: cannot concatenate 'str' and 'int' objects I'm not sure what to make of the above. Any help to get this working is greatly appreciated. -D From y.yaegashi at jp.fujitsu.com Wed Mar 7 01:13:52 2007 From: y.yaegashi at jp.fujitsu.com (Yoshihiko Yaegashi) Date: Wed, 07 Mar 2007 10:13:52 +0900 Subject: [et-mgmt-tools] About a solution to operate the virt-manager In-Reply-To: <20070306121553.GB21807@redhat.com> References: <45ED1D7E.9070805@jp.fujitsu.com> <20070306121553.GB21807@redhat.com> Message-ID: <45EE11D0.6070207@jp.fujitsu.com> Daniel P. Berrange wrote: > This is a GNOME / GTK bug I believe - try disabling 'Accessibility' > support in GNOME, and the logging out from X and strting a new > GNOME desktop session. Thank you for your advice. I have been possible to operate the virt-manager by the above method. I have a question. Is it possible to execute the Dogtail tests as 'Accessibility' support made disable? Thanks Yaegashi From berrange at redhat.com Wed Mar 7 01:58:45 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 7 Mar 2007 01:58:45 +0000 Subject: [et-mgmt-tools] About a solution to operate the virt-manager In-Reply-To: <45EE11D0.6070207@jp.fujitsu.com> References: <45ED1D7E.9070805@jp.fujitsu.com> <20070306121553.GB21807@redhat.com> <45EE11D0.6070207@jp.fujitsu.com> Message-ID: <20070307015845.GA9874@redhat.com> On Wed, Mar 07, 2007 at 10:13:52AM +0900, Yoshihiko Yaegashi wrote: > Daniel P. Berrange wrote: > > This is a GNOME / GTK bug I believe - try disabling 'Accessibility' > > support in GNOME, and the logging out from X and strting a new > > GNOME desktop session. > > Thank you for your advice. > I have been possible to operate the virt-manager by the above method. > > I have a question. > Is it possible to execute the Dogtail tests as 'Accessibility' > support made disable? No, it is not possible - Dogtail uses the accessibility extensions to control the application. Probably best to file a bug about the hang so we can make sure it gets tracked & resolved in Fedora 7. In the meantime things ought to be working correctly on FC6 with accessibility enabled. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From y.yaegashi at jp.fujitsu.com Wed Mar 7 05:26:08 2007 From: y.yaegashi at jp.fujitsu.com (Yoshihiko Yaegashi) Date: Wed, 07 Mar 2007 14:26:08 +0900 Subject: [et-mgmt-tools] About the install of PV domain by the virt-install Message-ID: <45EE4CF0.7040704@jp.fujitsu.com> Hi, I cannot install the Fedora 7 test2 of PV domain by the graphical mode of virt-install. (xen3.0.4, architecture:x86_64) Install has stoped on screen of attachment file. Thanks Yaegashi -------------- next part -------------- A non-text attachment was scrubbed... Name: fc7test2.png Type: image/png Size: 21842 bytes Desc: not available URL: From y.yaegashi at jp.fujitsu.com Wed Mar 7 06:17:58 2007 From: y.yaegashi at jp.fujitsu.com (Yoshihiko Yaegashi) Date: Wed, 07 Mar 2007 15:17:58 +0900 Subject: [et-mgmt-tools] About the install of PV domain by the virt-install In-Reply-To: <45EE4CF0.7040704@jp.fujitsu.com> References: <45EE4CF0.7040704@jp.fujitsu.com> Message-ID: <45EE5916.5040709@jp.fujitsu.com> Hi, The install of graphical mode by the VNC client connection fails under the following conditions. - architecture:x86_64 - PV domain (f7test2) - option '--nographics' - select the 'Start VNC' on the 'Unable to Start X' screen Install has stoped on screen of attachment file. Thanks Yaegashi -------------- next part -------------- A non-text attachment was scrubbed... Name: install.jpg Type: image/jpeg Size: 54667 bytes Desc: not available URL: From y.yaegashi at jp.fujitsu.com Wed Mar 7 06:42:11 2007 From: y.yaegashi at jp.fujitsu.com (Yoshihiko Yaegashi) Date: Wed, 07 Mar 2007 15:42:11 +0900 Subject: [et-mgmt-tools] About the install of PV domain by the virt-install In-Reply-To: <45EE4CF0.7040704@jp.fujitsu.com> References: <45EE4CF0.7040704@jp.fujitsu.com> Message-ID: <45EE5EC3.7000503@jp.fujitsu.com> Hi, The install of text mode by the virt-install fails under the following conditions. - architecture:x86_64 - PV domain (f7test2) - option '--nographics' - language choice:Japanese Install has stoped on screen of attachment file. Thanks Yaegashi -------------- next part -------------- A non-text attachment was scrubbed... Name: language.png Type: image/png Size: 39958 bytes Desc: not available URL: From fj0873gn at aa.jp.fujitsu.com Wed Mar 7 09:36:20 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Wed, 7 Mar 2007 18:36:20 +0900 Subject: [et-mgmt-tools] [PATCH] connecting console after install completes Message-ID: <200703071836.BHE35429.K9O840GH@aa.jp.fujitsu.com> Hi When the domain restarts after install completes, the console is not connected. Therefore, if I don't connect the console for myself, I cannot be finished the setting after install completes. The attached patch adds to solve this problem. Signed-off-by: Nobuhiro Itou Thanks, Nobuhiro Itou. Index: virt-install =================================================================== diff -r 3e18fa0cafc4 virt-install --- a/virt-install Fri Mar 02 09:16:33 2007 -0500 +++ b/virt-install Wed Mar 07 17:09:02 2007 +0900 @@ -476,6 +476,7 @@ def main(): # the domain print "Guest installation complete... restarting guest." dom.create() + guest.connect_console(conscb) else: print ("Domain installation does not appear to have been\n" "successful. If it was, you can restart your domain\n" Index: virtinst/Guest.py =================================================================== diff -r 3e18fa0cafc4 virtinst/Guest.py --- a/virtinst/Guest.py Fri Mar 02 09:16:33 2007 -0500 +++ b/virtinst/Guest.py Wed Mar 07 17:09:45 2007 +0900 @@ -547,6 +547,18 @@ class Guest(object): # for inactive guest, or get the still running install.. return self.conn.lookupByName(self.name) + def connect_console(self, consolecb): + child = None + if consolecb: + logging.debug("Launching console callback") + child = consolecb(self.domain) + + if child: # if we connected the console, wait for it to finish + try: + (pid, status) = os.waitpid(child, 0) + except OSError, (errno, msg): + print __name__, "waitpid:", msg + def validate_parms(self): if self.domain is not None: raise RuntimeError, "Domain already started!" From fj7716hz at aa.jp.fujitsu.com Wed Mar 7 09:45:26 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Wed, 07 Mar 2007 18:45:26 +0900 Subject: [et-mgmt-tools] [PATCH] Check the making domain's mac address Message-ID: <45EE89B6.5010200@aa.jp.fujitsu.com> Hi, The virt-install command can specify the making domain's vnif MAC address. The MAC address must be unique on the system, but the virt-install command doesn't check that the MAC address is unique among the running domains and host. The attached patch resolve this issue in the following way: 1) Get the running Domain's vnif MAC address. 2) Get the host's NIC MAC address. 3) Check the making domain's MAC address with 1) and 2) 's data. Signed-off-by: Tatsuro Enokura Thanks, Tatsuro Enokura. -------------------------------------------------------------------- diff -r 3e18fa0cafc4 virtinst/Guest.py --- a/virtinst/Guest.py Fri Mar 02 09:16:33 2007 -0500 +++ b/virtinst/Guest.py Wed Mar 07 18:34:59 2007 +0900 @@ -148,9 +148,40 @@ class VirtualNetworkInterface: self.macaddr = macaddr self.bridge = bridge - def setup(self): + def setup(self, conn): + # get the running domain's vNIC MACaddresss + macaddrlist = [] + ids = conn.listDomainsID(); + for id in ids: + vm = conn.lookupByID(id) + disks = util.get_network_devices(vm) + for (dummy, dummy, dummy, vm_macaddr) in disks: + macaddrlist.append(vm_macaddr) + + # get the Host's NIC MACaddress + host_macaddrlist = [] + hostdevs = util.get_host_network_devices() + for(dummy, dummy, dummy, dummy, host_macaddr) in hostdevs: + host_macaddrlist.append(host_macaddr) + if self.macaddr is None: - self.macaddr = util.randomMAC() + # macaddrs used check + retryFlg = True + while retryFlg: + self.macaddr = util.randomMAC() + for macaddr in macaddrlist: + if self.macaddr.upper() == macaddr.upper(): + break + else: + retryFlg = False + else: + for macaddr in macaddrlist: + if self.macaddr.upper() == macaddr.upper(): + raise ValueError, "MAC address has been already used by the running domain." + for macaddr in host_macaddrlist: + if self.macaddr.upper() == macaddr.upper(): + raise ValueError, "MAC address has been collided with the physical NIC." + if not self.bridge: self.bridge = util.default_bridge() @@ -401,7 +432,7 @@ class Guest(object): for disk in self.disks: disk.setup(progresscb) for nic in self.nics: - nic.setup() + nic.setup(self.conn) def _get_disk_xml(self, install = True): """Get the disk config in the libvirt XML format""" diff -r 3e18fa0cafc4 virtinst/util.py --- a/virtinst/util.py Fri Mar 02 09:16:33 2007 -0500 +++ b/virtinst/util.py Wed Mar 07 18:08:09 2007 +0900 @@ -14,6 +14,7 @@ import random import os.path +import libxml2 from sys import stderr def default_bridge(): @@ -123,3 +124,60 @@ def uuidFromString(s): def uuidFromString(s): s = s.replace('-', '') return [ int(s[i : i + 2], 16) for i in range(0, 32, 2) ] + +# the following function is from virt-manager/src/virtManager/domain.py +def get_network_devices(vm): + xml = vm.XMLDesc(0) + doc = None + try: + doc = libxml2.parseDoc(xml) + except: + return [] + ctx = doc.xpathNewContext() + disks = [] + try: + ret = ctx.xpathEval("/domain/devices/interface") + + for node in ret: + type = node.prop("type") + devmac = None + source = None + for child in node.children: + if child.name == "source": + if type == "bridge": + source = child.prop("bridge") + elif child.name == "mac": + devmac = child.prop("address") + + if source == None: + source = "-" + + devdst = "eth%d" % len(disks) + + disks.append([type, source, devdst, devmac]) + finally: + if ctx != None: + ctx.xpathFreeContext() + if doc != None: + doc.freeDoc() + return disks + +# the following function quotes from python2.5/uuid.py +def get_host_network_devices(): + device = [] + for dir in ['', '/sbin/', '/usr/sbin']: + executable = os.path.join(dir, "ifconfig") + if not os.path.exists(executable): + continue + try: + cmd = 'LC_ALL=C %s -a 2>/dev/null' % (executable) + pipe = os.popen(cmd) + except IOError: + continue + for line in pipe: + words = line.lower().split() + for i in range(len(words)): + if words[i] == "hwaddr": + device.append(words) + return device + -------------------------------------------------------------------- From dhawal at netmagicsolutions.com Wed Mar 7 11:01:26 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Wed, 07 Mar 2007 16:31:26 +0530 Subject: [et-mgmt-tools] Using cobbler to mirror repos like updates/extras -- further information In-Reply-To: <3cc2cbc40703061509j5051f6dgd15afa80eba22c51@mail.gmail.com> References: <45DF1A5F.10501@redhat.com> <20070223183317.GB3332@sliderule.msquared.com.au> <45DF3D12.4040006@redhat.com> <3cc2cbc40703061147u2dc922d6h8f9ac638044cfa98@mail.gmail.com> <45EDC97F.4090806@redhat.com> <3cc2cbc40703061232o59978a02u35f157cad9e1469f@mail.gmail.com> <45EDE0BC.8020207@redhat.com> <3cc2cbc40703061509j5051f6dgd15afa80eba22c51@mail.gmail.com> Message-ID: <45EE9B86.8070704@netmagicsolutions.com> Demetri Mouratis wrote: > Hi Michael, > > I went with the repo mirroring route: > > [root at hemops2 ~]# cobbler import --name=centos-4.4 > --mirror=rsync://rsync.mirrorservice.org/mirror.centos.org/4.4/os/i386/CentOS/ > > > This went fine. > > Adding the distro, however, gives me some trouble: > > [root at hemops2 ~]# cobbler distro add --name=centos4.4 > --kernel=/var/www/cobbler/repo_mirror/centos-4.4-i386/images/pxeboot/vmlinuz > > --initrd=/var/www/cobbler/repo_mirror/centos-4.4-i386/images/pxeboot/initrd.img This is resolved in 0.4.3, what version are you running? From dmourati at gmail.com Wed Mar 7 16:56:16 2007 From: dmourati at gmail.com (Demetri Mouratis) Date: Wed, 7 Mar 2007 08:56:16 -0800 Subject: [et-mgmt-tools] Using cobbler to mirror repos like updates/extras -- further information In-Reply-To: <45EE9B86.8070704@netmagicsolutions.com> References: <45DF1A5F.10501@redhat.com> <20070223183317.GB3332@sliderule.msquared.com.au> <45DF3D12.4040006@redhat.com> <3cc2cbc40703061147u2dc922d6h8f9ac638044cfa98@mail.gmail.com> <45EDC97F.4090806@redhat.com> <3cc2cbc40703061232o59978a02u35f157cad9e1469f@mail.gmail.com> <45EDE0BC.8020207@redhat.com> <3cc2cbc40703061509j5051f6dgd15afa80eba22c51@mail.gmail.com> <45EE9B86.8070704@netmagicsolutions.com> Message-ID: <3cc2cbc40703070856m4512aa2end1eeba698c80452f@mail.gmail.com> On 3/7/07, Dhawal Doshy wrote: > Demetri Mouratis wrote: > > Hi Michael, > > > > I went with the repo mirroring route: > > > > [root at hemops2 ~]# cobbler import --name=centos-4.4 > > --mirror=rsync://rsync.mirrorservice.org/mirror.centos.org/4.4/os/i386/CentOS/ > > > > > > This went fine. > > > > Adding the distro, however, gives me some trouble: > > > > [root at hemops2 ~]# cobbler distro add --name=centos4.4 > > --kernel=/var/www/cobbler/repo_mirror/centos-4.4-i386/images/pxeboot/vmlinuz > > > > --initrd=/var/www/cobbler/repo_mirror/centos-4.4-i386/images/pxeboot/initrd.img > > This is resolved in 0.4.3, what version are you running? > Hi Dhawal, I was on 0.4.2. An upgrade to 0.4.3 fixed the problem: [root at hemops2 ~]# rpm -Uvh /usr/src/redhat/RPMS/noarch/cobbler-0.4.3-0.noarch.rpm Preparing... ########################################### [100%] 1:cobbler ########################################### [100%] [root at hemops2 ~]# cobbler distro add --name=centos44 --kernel=/var/www/cobbler/repo_mirror/centos-4.4-i386/images/pxeboot/vmlinuz --initrd=/var/www/cobbler/repo_mirror/centos-4.4-i386/images/pxeboot/initrd.img [root at hemops2 ~]# cobbler list Distros: centos44 Repos: centos-4.4-i386 Profiles: Systems: Thanks. From mdehaan at redhat.com Wed Mar 7 21:18:07 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 07 Mar 2007 16:18:07 -0500 Subject: Xen and Fully Automatic Kickstarts (Was: [et-mgmt-tools] More Cobbler, Koan and Xen questions) In-Reply-To: <514B41B4EB0E834EAFB17FA73E263B7009B32D@exch001.ibsys.com> References: <514B41B4EB0E834EAFB17FA73E263B7009B32D@exch001.ibsys.com> Message-ID: <45EF2C0E.10504@redhat.com> Tim Foreman asked yesterday kickstarts appear to not be used with Xen and "koan --virt". I found something today that might help explain that problem. Usually I use IP addresses in /var/lib/cobbler/settings, but today I was testing Xen and using FQDNs. I am not entirely sure if this helps in Tim's case, but at least on my system, if I use hostnames in /var/lib/cobbler/settings for the server parameter, Xen will not retrieve the kickstart -- however if I use an IP, it does, and it's fully automatic. (This is on FC6). So, if you have hostnames in /var/lib/cobbler/settings, changing those values to IPs and then running "sync" may help with Xen kickstart issues. --Michael From berrange at redhat.com Thu Mar 8 00:16:16 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 8 Mar 2007 00:16:16 +0000 Subject: [et-mgmt-tools] About a solution to operate the virt-manager In-Reply-To: <20070307015845.GA9874@redhat.com> References: <45ED1D7E.9070805@jp.fujitsu.com> <20070306121553.GB21807@redhat.com> <45EE11D0.6070207@jp.fujitsu.com> <20070307015845.GA9874@redhat.com> Message-ID: <20070308001616.GZ6302@redhat.com> On Wed, Mar 07, 2007 at 01:58:45AM +0000, Daniel P. Berrange wrote: > On Wed, Mar 07, 2007 at 10:13:52AM +0900, Yoshihiko Yaegashi wrote: > > Daniel P. Berrange wrote: > > > This is a GNOME / GTK bug I believe - try disabling 'Accessibility' > > > support in GNOME, and the logging out from X and strting a new > > > GNOME desktop session. > > > > Thank you for your advice. > > I have been possible to operate the virt-manager by the above method. > > > > I have a question. > > Is it possible to execute the Dogtail tests as 'Accessibility' > > support made disable? > > No, it is not possible - Dogtail uses the accessibility extensions to > control the application. Probably best to file a bug about the hang so > we can make sure it gets tracked & resolved in Fedora 7. In the meantime > things ought to be working correctly on FC6 with accessibility enabled. I've just confirmed the hang and identified the flaw in PyGTK bindings that causes it. The BZ is https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=231394 If this doesn't get addressed promptly I'll see if I can write a small C module to workaround the flaw by forcably calling gdk_threads_init() Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From dmourati at gmail.com Thu Mar 8 01:59:42 2007 From: dmourati at gmail.com (Demetri Mouratis) Date: Wed, 7 Mar 2007 17:59:42 -0800 Subject: [et-mgmt-tools] Two Minor Problems with Cobbler Message-ID: <3cc2cbc40703071759s5a60ea59s6cde47ffb8fe5c5f@mail.gmail.com> Hi, I've been spending quite a bit of time with cobbler lately and have run into two very small issues. First, running cobbler sync with manage_dhcp: 1 set drops an empty dhcpd.conf directory in the PWD. [root at hemops2 ~]# cd /tmp/ [root at hemops2 tmp]# ls [root at hemops2 tmp]# cobbler sync sync distro: centos44 Shutting down dhcpd: [ OK ] Starting dhcpd: [ OK ] [root at hemops2 tmp]# ls dhcpd.conf Second, I have two configuration changes I need made to /tftpboot/pxelinux.cfg/default. These have to do with NIC ordering and a doc I found here: http://linux.dell.com/files/whitepapers/nic-enum-whitepaper-v2.pdf There, it suggests making two changes to the pxe configuration, adding the following lines: IPAPPEND 2 APPEND ksdevice=bootif I put a config file called default.pxe in /etc/cobbler and run cobbler sync but I see a vanilla default file in /tftpboot/pxelinux.cfg/default. Thanks. From takatom at jp.fujitsu.com Thu Mar 8 08:27:53 2007 From: takatom at jp.fujitsu.com (Takahashi Tomohiro) Date: Thu, 8 Mar 2007 17:27:53 +0900 Subject: [et-mgmt-tools] [PATCH][RFC] Add --keymap option to virt-install Message-ID: <018301c7615b$ae8c58a0$ecb1220a@fj98229700> Hi, Dan Thank you for your comments. Please commit this keymap patch, if it hasn't problem. Thanks, Tomohiro Takahashi. > On Tue, Mar 06, 2007 at 09:59:16PM +0900, Takahashi Tomohiro wrote: >> Hi, >> >> I would like to add --keymap option to virt-install in order to set up >> jp106 keymap. Because I can't use jp106 keyboard. So, I made a patch >> (notes:It relates libvirt) >> >> Usage: virt-install --keymap=ja >> >> I confirmed that I can use jp106 keyboard on HVM domain. >> But didn't confirm PV domain. >> >> Please give me an advice, if you have it. > > ACK. The patch looks fine to me. > > Regards, > Dan. > -- > |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 > 2496 -=| > |=- Perl modules: > .cpan.org/~danberr/ -=| > |=- Projects: > at.net/~danielpb/ -=| > |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B > 505 -=| > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > From dhawal at netmagicsolutions.com Thu Mar 8 09:16:15 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Thu, 08 Mar 2007 14:46:15 +0530 Subject: [et-mgmt-tools] Two Minor Problems with Cobbler In-Reply-To: <3cc2cbc40703071759s5a60ea59s6cde47ffb8fe5c5f@mail.gmail.com> References: <3cc2cbc40703071759s5a60ea59s6cde47ffb8fe5c5f@mail.gmail.com> Message-ID: <45EFD45F.2010907@netmagicsolutions.com> Demetri Mouratis wrote: > Hi, > > I've been spending quite a bit of time with cobbler lately and have > run into two very small issues. First, running cobbler sync with > manage_dhcp: 1 set drops an empty dhcpd.conf directory in the PWD. > > [root at hemops2 ~]# cd /tmp/ > [root at hemops2 tmp]# ls > [root at hemops2 tmp]# cobbler sync > sync distro: centos44 > Shutting down dhcpd: [ OK ] > Starting dhcpd: [ OK ] > [root at hemops2 tmp]# ls > dhcpd.conf i can confirm presence of 2 empty directories under PWD post 'cobbler sync' in my setup as well. dhcpd.conf ks.cfg > Second, I have two configuration changes I need made to > /tftpboot/pxelinux.cfg/default. These have to do with NIC ordering and > a doc I found here: > > http://linux.dell.com/files/whitepapers/nic-enum-whitepaper-v2.pdf > > There, it suggests making two changes to the pxe configuration, adding > the following lines: > > IPAPPEND 2 > APPEND ksdevice=bootif You can also use an undocumented 'ksdevice: link' in your /var/lib/cobbler/settings under 'kernel_options:' - dhawal From jeevanullas at gmail.com Thu Mar 8 11:40:46 2007 From: jeevanullas at gmail.com (Deependra Shekhawat) Date: Thu, 8 Mar 2007 17:10:46 +0530 Subject: [et-mgmt-tools] newbie to the term cobbler Message-ID: <57127b5f0703080340u3c193100r37305acb0af63cb2@mail.gmail.com> Hello guys, I got through this term "cobbler" often in this mailing list. I saw it's website. There it says cobbler does provisioning. now what exactly does provisioning means? what cobbler is? please help me with these basic questions and let me know where to start off if i want to learn about it more. regards Deependra -- Enjoy Life ! -------------- next part -------------- An HTML attachment was scrubbed... URL: From takatom at jp.fujitsu.com Thu Mar 8 12:14:19 2007 From: takatom at jp.fujitsu.com (Takahashi Tomohiro) Date: Thu, 8 Mar 2007 21:14:19 +0900 Subject: [et-mgmt-tools] [PATCH][RFC] Add --keymap option to virt-install References: <018301c7615b$ae8c58a0$ecb1220a@fj98229700> Message-ID: <021501c7617b$5065a9a0$ecb1220a@fj98229700> Hi, Previously I tested keymap patch on virt-inst which was already posted. Today I test the patch on virt-manager.Then I found a bug. This patch fixes this problem. I would like to post an update patch. Because I caught a bug in keymap patch that I had posted on March 6. Errors did happen, when I used virt-manager. Because It doesn't have a function for keymap. virtmanager does not have keymap setting function. So I add a check "keymap" exists or not. Please refer to following error message. Traceback (most recent call last): File "/usr/local/share/virt-manager/virtManager/create.py", line 508, in finish guest.graphics = "vnc" File "/usr/lib/python2.5/site-packages/virtinst/Guest.py", line 414, in set_graphics gt = VNCVirtualGraphics(opts, self._graphics["keymap"]) KeyError: 'keymap' The difference is following in virtinst/Guest.py. modification: ===FIXED==== @@ -386,7 +411,10 @@ class Guest(object): if self._graphics["enabled"] == True: if t == "vnc": - gt = VNCVirtualGraphics(opts) + if self.graphics.has_key("keymap"): + gt = VNCVirtualGraphics(opts, self._graphics["keymap"]) + else: + gt = VNCVirtualGraphics(opts) elif t == "sdl": gt = SDLVirtualGraphics(opts) else: ===BEFORE=== @@ -386,7 +411,7 @@ class Guest(object): if self._graphics["enabled"] == True: if t == "vnc": - gt = VNCVirtualGraphics(opts) + gt = VNCVirtualGraphics(opts, self._graphics["keymap"]) elif t == "sdl": gt = SDLVirtualGraphics(opts) else: ========== Signed-off-by: Tomohiro Takahashi Thanks, Tomohiro Takahashi. =================================================================== diff -uNrp virtinst-0.101.0.orig/virt-install virtinst-0.101.0/virt-install --- virtinst-0.101.0.orig/virt-install 2007-03-05 21:29:44.000000000 +0900 +++ virtinst-0.101.0/virt-install 2007-03-06 15:51:22.000000000 +0900 @@ -96,6 +96,13 @@ def get_vcpus(vcpus, guest): except ValueError, e: print "ERROR: ", e +def get_keymap(keymap, guest): + if keymap: + try: + guest.keymap = keymap + except ValueError, e: + print "ERROR: ", e + def get_disk(disk, size, sparse, guest, hvm): # FIXME: need to handle a list of disks at some point while 1: @@ -160,14 +167,14 @@ def get_networks(macs, bridges, guest): else: get_network(macs, bridges, guest) -def get_graphics(vnc, vncport, nographics, sdl, guest): +def get_graphics(vnc, vncport, nographics, sdl, keymap, guest): if vnc and nographics: raise ValueError, "Can't do both VNC graphics and nographics" if nographics: guest.graphics = False return if vnc is not None: - guest.graphics = (True, "vnc", vncport) + guest.graphics = (True, "vnc", vncport, keymap) return if sdl is not None: guest.graphics = (True, "sdl") @@ -180,7 +187,7 @@ def get_graphics(vnc, vncport, nographic print "ERROR", e continue if vnc: - guest.graphics = "vnc" + guest.graphics = (True, "vnc", vncport, keymap) else: guest.graphics = False break @@ -256,6 +263,9 @@ def parse_args(): action="store_false", dest="autoconsole", help="Don't automatically try to connect to the guest console") + parser.add_option("-k", "--keymap", type="string", dest="keymap", + help="set up keymap for a graphical console") + parser.add_option("", "--accelerate", action="store_true", dest="accelerate", help="Use kernel acceleration capabilities") parser.add_option("", "--connect", type="string", dest="connect", @@ -417,6 +427,7 @@ def main(): get_memory(options.memory, guest) get_uuid(options.uuid, guest) get_vcpus(options.vcpus, guest) + get_keymap(options.keymap, guest) # set up disks get_disks(options.diskfile, options.disksize, options.sparse, @@ -426,7 +437,7 @@ def main(): get_networks(options.mac, options.bridge, guest) # set up graphics information - get_graphics(options.vnc, options.vncport, options.nographics, options.sdl, guest) + get_graphics(options.vnc, options.vncport, options.nographics, options.sdl, options.keymap, guest) # and now for the full-virt vs paravirt specific questions if not hvm: # paravirt diff -uNrp virtinst-0.101.0.orig/virtinst/Guest.py virtinst-0.101.0/virtinst/Guest.py --- virtinst-0.101.0.orig/virtinst/Guest.py 2007-03-05 21:29:44.000000000 +0900 +++ virtinst-0.101.0/virtinst/Guest.py 2007-03-08 20:01:04.000000000 +0900 @@ -185,9 +185,17 @@ class VNCVirtualGraphics(XenGraphics): self.port = args[0] else: self.port = -1 + if len(args) >= 2 and args[1]: + self.keymap = args[1] + else: + self.keymap = None def get_xml_config(self): - return " " % (self.port) + if self.keymap == None: + keymapstr = "" + else: + keymapstr = "keymap='"+self.keymap+"' " + return " " % {"port":self.port, "keymapstr":keymapstr} # Back compat class to avoid ABI break class XenVNCGraphics(VNCVirtualGraphics): @@ -218,6 +226,7 @@ class Guest(object): self._maxmemory = None self._vcpus = None self._graphics = { "enabled": False } + self._keymap = None self.domain = None self.conn = connection @@ -301,6 +310,21 @@ class Guest(object): self._vcpus = val vcpus = property(get_vcpus, set_vcpus) + + # keymap for the guest + def get_keymap(self): + return self._keymap + def set_keymap(self, val): + if val and (re.match("^[a-zA-Z0-9_]*$", val) == None): + raise ValueError, "Keymap be alphanumeric or _" + if val and (len(val) > 16): + raise ValueError, "Keymap must be less than 16 characters" + if val and (type(val) != type("string")): + raise ValueError, "Keymap must be a string" + self._keymap = val + keymap = property(get_keymap, set_keymap) + + # kernel + initrd pair to use for installing as opposed to using a location def get_boot(self): return self._boot @@ -374,6 +398,7 @@ class Guest(object): if len(val) >= 1: self._graphics["enabled"] = val[0] if len(val) >= 2: t = val[1] if len(val) >= 3: opts = val[2] + if len(val) >= 4: self._graphics["keymap"] = val[3] else: if val in ("vnc", "sdl"): t = val @@ -386,7 +411,10 @@ class Guest(object): if self._graphics["enabled"] == True: if t == "vnc": - gt = VNCVirtualGraphics(opts) + if self.graphics.has_key("keymap"): + gt = VNCVirtualGraphics(opts, self._graphics["keymap"]) + else: + gt = VNCVirtualGraphics(opts) elif t == "sdl": gt = SDLVirtualGraphics(opts) else: > Hi, Dan > > Thank you for your comments. > Please commit this keymap patch, if it hasn't problem. > > Thanks, > Tomohiro Takahashi. > >> On Tue, Mar 06, 2007 at 09:59:16PM +0900, Takahashi Tomohiro wrote: >>> Hi, >>> >>> I would like to add --keymap option to virt-install in order to set up >>> jp106 keymap. Because I can't use jp106 keyboard. So, I made a patch >>> (notes:It relates libvirt) >>> >>> Usage: virt-install --keymap=ja >>> >>> I confirmed that I can use jp106 keyboard on HVM domain. >>> But didn't confirm PV domain. >>> >>> Please give me an advice, if you have it. >> >> ACK. The patch looks fine to me. >> >> Regards, >> Dan. >> -- >> |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 >> 2496 -=| >> |=- Perl modules: .cpan.org/~danberr/ -=| >> |=- Projects: at.net/~danielpb/ -=| >> |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B >> 05 -=| >> > From berrange at redhat.com Thu Mar 8 12:46:23 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 8 Mar 2007 12:46:23 +0000 Subject: [et-mgmt-tools] [PATCH][RFC] Add --keymap option to virt-install In-Reply-To: <021501c7617b$5065a9a0$ecb1220a@fj98229700> References: <018301c7615b$ae8c58a0$ecb1220a@fj98229700> <021501c7617b$5065a9a0$ecb1220a@fj98229700> Message-ID: <20070308124623.GC9007@redhat.com> On Thu, Mar 08, 2007 at 09:14:19PM +0900, Takahashi Tomohiro wrote: > Hi, > > Previously I tested keymap patch on virt-inst which was already posted. > Today I test the patch on virt-manager.Then I found a bug. > This patch fixes this problem. > > I would like to post an update patch. Because I caught a bug in keymap patch > that I had posted on March 6. Errors did happen, when I used virt-manager. > Because It doesn't have a function for keymap. Hmm, now that we have this virt-install patch, we'll need some way to let people set the keymap from virt-manager too. I imagine we could add some user preference to the 'Preferences' dialog box, either a plain text box where they can enter a keymap, or perhaps a drop down list pre-populated with all keymaps. Unless there is some way to automatically pick the correct keymap based on the X keymap - but i don't think we get get at that info any way. > virtmanager does not have keymap setting function. > So I add a check "keymap" exists or not. Applied this updated patch to virt-install HG repo now.. http://hg.et.redhat.com/virt/applications/virtinst--devel?cs=660a6db5a4be Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From mark.cave-ayland at ilande.co.uk Thu Mar 8 14:32:49 2007 From: mark.cave-ayland at ilande.co.uk (Mark Cave-Ayland) Date: Thu, 08 Mar 2007 14:32:49 +0000 Subject: [et-mgmt-tools] virt-manager right-click vm options patch Message-ID: <1173364369.13799.5.camel@mca-desktop> Hi everyone, Please find enclosed a patch that allows the user to start, stop, pause and resume VMs by adding these options to the right-click menu when selecting a VM. Note that this is my first patch involving all these technologies (virt-manager, libvirt and python) so it will require a reasonable review. Many thanks, Mark. -------------- next part -------------- A non-text attachment was scrubbed... Name: virt-manager-rightclick.patch Type: text/x-patch Size: 5830 bytes Desc: not available URL: From hbrock at redhat.com Thu Mar 8 15:12:00 2007 From: hbrock at redhat.com (Hugh Brock) Date: Thu, 08 Mar 2007 10:12:00 -0500 Subject: [et-mgmt-tools] virt-manager right-click vm options patch In-Reply-To: <1173364369.13799.5.camel@mca-desktop> References: <1173364369.13799.5.camel@mca-desktop> Message-ID: <45F027C0.4070202@redhat.com> Mark Cave-Ayland wrote: > Hi everyone, > > Please find enclosed a patch that allows the user to start, stop, pause > and resume VMs by adding these options to the right-click menu when > selecting a VM. Note that this is my first patch involving all these > technologies (virt-manager, libvirt and python) so it will require a > reasonable review. > > > Many thanks, > > Mark. > Hi, this looks good, I've been meaning to do this myself for a while. I'm testing it now; if it works, I'll apply it. --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From mdehaan at redhat.com Thu Mar 8 16:48:48 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 08 Mar 2007 11:48:48 -0500 Subject: [et-mgmt-tools] Two Minor Problems with Cobbler In-Reply-To: <3cc2cbc40703071759s5a60ea59s6cde47ffb8fe5c5f@mail.gmail.com> References: <3cc2cbc40703071759s5a60ea59s6cde47ffb8fe5c5f@mail.gmail.com> Message-ID: <45F03E70.9010804@redhat.com> Demetri Mouratis wrote: > Hi, > > I've been spending quite a bit of time with cobbler lately and have > run into two very small issues. First, running cobbler sync with > manage_dhcp: 1 set drops an empty dhcpd.conf directory in the PWD. Good catch... this apparently came across in the move back to Cheetah for templating. I'll commit and push the fix. [mdehaan at mdehaan cobbler]$ hg diff diff -r 7f3dc5cdd1b40c7521d2ce2052388e7df69906dc cobbler/action_sync.py --- a/cobbler/action_sync.py Fri Mar 02 11:07:10 2007 -0500 +++ b/cobbler/action_sync.py Thu Mar 08 11:40:52 2007 -0500 @@ -464,7 +464,7 @@ class BootSync: t = Template(source=data, searchList=[metadata]) data_out = str(t) - self.mkdir(os.path.basename(out_path)) + self.mkdir(os.path.dirname(out_path)) fd = open(out_path, "w+") fd.write(data_out) fd.close() > > [root at hemops2 ~]# cd /tmp/ > [root at hemops2 tmp]# ls > [root at hemops2 tmp]# cobbler sync > sync distro: centos44 > Shutting down dhcpd: [ OK ] > Starting dhcpd: [ OK ] > [root at hemops2 tmp]# ls > dhcpd.conf > > Second, I have two configuration changes I need made to > /tftpboot/pxelinux.cfg/default. These have to do with NIC ordering and > a doc I found here: > > http://linux.dell.com/files/whitepapers/nic-enum-whitepaper-v2.pdf > > There, it suggests making two changes to the pxe configuration, adding > the following lines: > > IPAPPEND 2 > APPEND ksdevice=bootif > I put a config file called default.pxe in /etc/cobbler and run cobbler > sync but I see a vanilla default file in > /tftpboot/pxelinux.cfg/default. Kickstart parameters can be managed with a cobbler profile or even on specific systems or distros. "cobbler profile add ... --kopts="ksdevice=bootif otheroption=blah etc etc foo etc" The vanilla cobbler "default" file in recent versions now is used for PXE menus. If you would like to use a specific profile as a default profile instead, you can do the following: cobbler system add --name=default --profile=name-of-profile-to-use-as-default If you have a kernel option you would like to use on every single cobbler profile you have, you can also add that kernel option to /var/lib/cobbler/settings and it will automagically appear on every profile and system. (run "cobbler sync" to apply changes) > > Thanks. > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From mdehaan at redhat.com Thu Mar 8 16:59:04 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 08 Mar 2007 11:59:04 -0500 Subject: [et-mgmt-tools] newbie to the term cobbler In-Reply-To: <57127b5f0703080340u3c193100r37305acb0af63cb2@mail.gmail.com> References: <57127b5f0703080340u3c193100r37305acb0af63cb2@mail.gmail.com> Message-ID: <45F040D8.1090501@redhat.com> Deependra Shekhawat wrote: > Hello guys, > > I got through this term "cobbler" often in this mailing list. I saw > it's website. There it says cobbler does provisioning. now what > exactly does provisioning means? In this context, provisioning refers to the automated (and usually remote) installation of operating systems and associated software. > what cobbler is? please help me with these basic questions and let me > know where to start off if i want to learn about it more. I'd start with the slides here -- http://cobbler.et.redhat.com/cobbler.odp -- Open Office Format. > > regards > Deependra > > -- > Enjoy Life ! > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From dmourati at gmail.com Thu Mar 8 17:03:43 2007 From: dmourati at gmail.com (Demetri Mouratis) Date: Thu, 8 Mar 2007 09:03:43 -0800 Subject: [et-mgmt-tools] Two Minor Problems with Cobbler In-Reply-To: <45F03E70.9010804@redhat.com> References: <3cc2cbc40703071759s5a60ea59s6cde47ffb8fe5c5f@mail.gmail.com> <45F03E70.9010804@redhat.com> Message-ID: <3cc2cbc40703080903k243e35edo27bdd66aec2df67a@mail.gmail.com> On 3/8/07, Michael DeHaan wrote: Hi Michael, Thanks for looking at the directory creation issue I found. > > > > Second, I have two configuration changes I need made to > > /tftpboot/pxelinux.cfg/default. These have to do with NIC ordering and > > a doc I found here: > > > > http://linux.dell.com/files/whitepapers/nic-enum-whitepaper-v2.pdf > > > > There, it suggests making two changes to the pxe configuration, adding > > the following lines: > > > > IPAPPEND 2 > > APPEND ksdevice=bootif > > > I put a config file called default.pxe in /etc/cobbler and run cobbler > > sync but I see a vanilla default file in > > /tftpboot/pxelinux.cfg/default. > > Kickstart parameters can be managed with a cobbler profile or even on > specific systems or distros. > > "cobbler profile add ... --kopts="ksdevice=bootif otheroption=blah etc > etc foo etc" > > The vanilla cobbler "default" file in recent versions now is used for > PXE menus. If you would like to use a specific profile as a default > profile instead, you can do the following: > > cobbler system add --name=default > --profile=name-of-profile-to-use-as-default > > If you have a kernel option you would like to use on every single > cobbler profile you have, you can also add that kernel option to > /var/lib/cobbler/settings and it will automagically appear on every > profile and system. (run "cobbler sync" to apply changes) In this case, I am concerned with a problem at PXE boot time, prior to kickstart. In particular, the IPAPPEND 2 line above, I thought the IPAPPEND was an element in the pxe configuration, not a kernel option. For ksdevice, yea, I can see how to do that now. My point was, I didn't see a good way to make changes to /tftboot/pxelinux.cfg/default stick. Maybe I misunderstood the docs. What is the point of /etc/cobbler/default.pxe? Thanks. From mdehaan at redhat.com Thu Mar 8 17:16:10 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 08 Mar 2007 12:16:10 -0500 Subject: [et-mgmt-tools] Two Minor Problems with Cobbler In-Reply-To: <3cc2cbc40703080903k243e35edo27bdd66aec2df67a@mail.gmail.com> References: <3cc2cbc40703071759s5a60ea59s6cde47ffb8fe5c5f@mail.gmail.com> <45F03E70.9010804@redhat.com> <3cc2cbc40703080903k243e35edo27bdd66aec2df67a@mail.gmail.com> Message-ID: <45F044DA.2000807@redhat.com> Demetri Mouratis wrote: > On 3/8/07, Michael DeHaan wrote: > > Hi Michael, > > Thanks for looking at the directory creation issue I found. > > >> > >> > Second, I have two configuration changes I need made to >> > /tftpboot/pxelinux.cfg/default. These have to do with NIC ordering and >> > a doc I found here: >> > >> > http://linux.dell.com/files/whitepapers/nic-enum-whitepaper-v2.pdf >> > >> > There, it suggests making two changes to the pxe configuration, adding >> > the following lines: >> > >> > IPAPPEND 2 >> > APPEND ksdevice=bootif >> >> > I put a config file called default.pxe in /etc/cobbler and run cobbler >> > sync but I see a vanilla default file in >> > /tftpboot/pxelinux.cfg/default. >> >> Kickstart parameters can be managed with a cobbler profile or even on >> specific systems or distros. >> >> "cobbler profile add ... --kopts="ksdevice=bootif otheroption=blah etc >> etc foo etc" >> >> The vanilla cobbler "default" file in recent versions now is used for >> PXE menus. If you would like to use a specific profile as a default >> profile instead, you can do the following: >> >> cobbler system add --name=default >> --profile=name-of-profile-to-use-as-default >> >> If you have a kernel option you would like to use on every single >> cobbler profile you have, you can also add that kernel option to >> /var/lib/cobbler/settings and it will automagically appear on every >> profile and system. (run "cobbler sync" to apply changes) > > In this case, I am concerned with a problem at PXE boot time, prior to > kickstart. In particular, the IPAPPEND 2 line above, I thought the > IPAPPEND was an element in the pxe configuration, not a kernel option. > For ksdevice, yea, I can see how to do that now. > > My point was, I didn't see a good way to make changes to > /tftboot/pxelinux.cfg/default stick. Maybe I misunderstood the docs. > What is the point of /etc/cobbler/default.pxe? > > Thanks. > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools /etc/cobbler/default.pxe is unmanaged by the RPM at this point -- as it was flagged a configuration file, RPM won't delete it when it's no longer managed. As for IPAPPEND2, I was not familiar with it, but the docs ... http://syslinux.zytor.com/faq.php#config ... seem to indicate that only affects kernel arguments. This is good. Cobbler can muck with kernel arguments on it's own, using system --kopts, but that's probably more work than you want if you are not using "cobbler system add" already. If you feel like beta testing, try this... [mdehaan at mdehaan cobbler]$ hg diff diff -r a0817e4a44f6e6b732a2ab350dfe2ca838a71f76 cobbler/action_sync.py --- a/cobbler/action_sync.py Thu Mar 08 12:00:16 2007 -0500 +++ b/cobbler/action_sync.py Thu Mar 08 12:12:46 2007 -0500 @@ -622,6 +622,7 @@ class BootSync: buffer = buffer + "prompt 0\n" buffer = buffer + "timeout 1\n" buffer = buffer + "label linux\n" + buffer = buffer + "ipappend 2\n" buffer = buffer + "\tkernel %s\n" % kernel_path else: # elilo thrown in root And then all profiles will use that. Does anyone see a problem with IPAPPEND 2 being added all of the time? It seems like it would cause no adverse affects. If not, I can make it a configuration option. --Michael From dhawal at netmagicsolutions.com Thu Mar 8 17:40:03 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Thu, 08 Mar 2007 23:10:03 +0530 Subject: [et-mgmt-tools] Low priority feature request.. (memtest distro) Message-ID: <45F04A73.7090400@netmagicsolutions.com> I am trying to add memtest as one of the distros.. # cobbler distro add --name=memtest --kernel=/tmp/memtest cannot find kernel file # file /tmp/memtest /tmp/memtest: x86 boot sector A 'real' kernel also lists as a 'x86 boot sector' file.. # file /var/www/cobbler/images/centos/vmlinuz /var/www/cobbler/images/centos/vmlinuz: x86 boot sector Can this be considered as a low priority feature request? - dhawal From mdehaan at redhat.com Thu Mar 8 17:59:26 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 08 Mar 2007 12:59:26 -0500 Subject: [et-mgmt-tools] Low priority feature request.. (memtest distro) In-Reply-To: <45F04A73.7090400@netmagicsolutions.com> References: <45F04A73.7090400@netmagicsolutions.com> Message-ID: <45F04EFE.7070302@redhat.com> Dhawal Doshy wrote: > I am trying to add memtest as one of the distros.. > # cobbler distro add --name=memtest --kernel=/tmp/memtest > cannot find kernel file > > # file /tmp/memtest > /tmp/memtest: x86 boot sector > > A 'real' kernel also lists as a 'x86 boot sector' file.. > # file /var/www/cobbler/images/centos/vmlinuz > /var/www/cobbler/images/centos/vmlinuz: x86 boot sector > > Can this be considered as a low priority feature request? > > - dhawal > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools The file type checking implementation in cobbler currently a bit "lame", and looks to see if something has "vmlinuz" in it for kernel names. Executing "file" for local files would be smarter. The one catch here is that there are two kinds of initrd files -- CPIO and gzips, so that is something to be careful about. This would involve some small changes to util.py. Another idea -- rather than having a cobbler "memtest" distro (which would require a lot of hacking to prevent files being written to /var/lib/cobbler and confusing koan), it might be better to have cobbler, when writing out the PXE menus, just to create an additional PXE menu entry for memtest, possibly adding it to the cobbler package if it's not clearly available in another package. This entry could appear last or first in the list, either way. It defintely wouldn't need to be a full fledged distro (with an entry in /var/lib/cobbler/distros, etc) as it's not really installable. This would require changes just in action_sync.py Both of these seem pretty easy to do ... perhaps someone on-list would like to submit a patch or two? Wink, wink, nudge, nudge? --Michael From hbrock at redhat.com Thu Mar 8 19:41:57 2007 From: hbrock at redhat.com (Hugh Brock) Date: Thu, 08 Mar 2007 14:41:57 -0500 Subject: [et-mgmt-tools] virt-manager right-click vm options patch In-Reply-To: <1173364369.13799.5.camel@mca-desktop> References: <1173364369.13799.5.camel@mca-desktop> Message-ID: <45F06705.5070809@redhat.com> Mark Cave-Ayland wrote: > Hi everyone, > > Please find enclosed a patch that allows the user to start, stop, pause > and resume VMs by adding these options to the right-click menu when > selecting a VM. Note that this is my first patch involving all these > technologies (virt-manager, libvirt and python) so it will require a > reasonable review. > > > Many thanks, > I've committed this. Thanks! -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From mdehaan at redhat.com Thu Mar 8 23:24:13 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 08 Mar 2007 18:24:13 -0500 Subject: [et-mgmt-tools] Greatly improved Xen support for koan Message-ID: <45F09B1D.2010007@redhat.com> For those using Cobbler for virtualization installs, I've commited a change to the Mercurial repository that you will probably be interested in. Basically koan now uses the "virtinst" library to install virtualized images rather than using "libvirt" directly. "virtinst" is the same library that virt-manager uses to install virtual machines. So, how does this affect the running of koan? -- The biggest change is that koan now terminates very quickly, with the rest of the virtual installation going on in the background. This allows multiple parallel virtual installs if someone wants to do it. This should make koan much more usable in batch context. If you want to watch the installation progress, usage of xm or virt-manager will allow that). -- Koan also names each virtual system after the mac address they are created with, such that they are unique. So no more "xenguest" and "xenguest_1" kind of thing, which was error prone when you installed a machine and then it crashed, causing naming conflicts. (--xen-name will probably be removed from cobbler also, as it's no longer needed). If you are using koan with "--system" instead of "--profile", the MAC address can be controlled from cobbler, along with other fine tunning arguments. If not, it will be randomly generated. If anyone has any objections to this naming scheme, you might want to speak up. I can think of a couple of alternate ways to name domains automatically -- "$profilename_$timestamp" isn't so bad. -- Koan no longer writes any files to /var/koan -- anywhere, because it doesn't have to. This is still undergoing some testing, but I would call it very stable at this point. I'm just going to be tweaking output and error handling a bit, if anything. Further down the road, this will enable koan to do hardware virtual installs and handle other virtualization types. Note that depending on what distro you are running, Xen will save the configuration files in one of the following: /etc/xen /var/lib/xen /var/run/xen So, if you have domains that get stopped, looking in those directories will tell you which ones you have to start. Also on FC6 and later you can experiment with "virsh" for management of inactive domains. 0.2.7 will probably be released sometime later next week ... unless you want to play with the upstream code (please do, not too many computers have exploded from using it, I promise) --Michael From berrange at redhat.com Thu Mar 8 23:29:46 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 8 Mar 2007 23:29:46 +0000 Subject: [et-mgmt-tools] Greatly improved Xen support for koan In-Reply-To: <45F09B1D.2010007@redhat.com> References: <45F09B1D.2010007@redhat.com> Message-ID: <20070308232945.GB10363@redhat.com> On Thu, Mar 08, 2007 at 06:24:13PM -0500, Michael DeHaan wrote: > For those using Cobbler for virtualization installs, I've commited a > change to the Mercurial repository that you will probably be interested > in. > > Basically koan now uses the "virtinst" library to install virtualized > images rather than using "libvirt" directly. "virtinst" is the same > library that virt-manager uses to install virtual machines. So, how > does this affect the running of koan? > > -- The biggest change is that koan now terminates very quickly, with the > rest of the virtual installation going on in the background. This > allows multiple parallel virtual installs if someone wants to do it. > This should make koan much more usable in batch context. If you want > to watch the installation progress, usage of xm or virt-manager will > allow that). > -- Koan also names each virtual system after the mac address they are > created with, such that they are unique. So no more "xenguest" and > "xenguest_1" kind of thing, which was error prone when you installed a > machine and then it crashed, causing naming conflicts. (--xen-name > will probably be removed from cobbler also, as it's no longer needed). > If you are using koan with "--system" instead of "--profile", the MAC > address can be controlled from cobbler, along with other fine tunning > arguments. If not, it will be randomly generated. If anyone has any > objections to this naming scheme, you might want to speak up. I can > think of a couple of alternate ways to name domains automatically -- > "$profilename_$timestamp" isn't so bad. > -- Koan no longer writes any files to /var/koan -- anywhere, because it > doesn't have to. > > This is still undergoing some testing, but I would call it very stable > at this point. I'm just going to be tweaking output and error handling > a bit, if anything. > Further down the road, this will enable koan to do hardware virtual > installs and handle other virtualization types. > > Note that depending on what distro you are running, Xen will save the > configuration files in one of the following: > /etc/xen > /var/lib/xen > /var/run/xen Scrub those last two - its /etc/xen or /var/lib/xend/domains > So, if you have domains that get stopped, looking in those directories > will tell you which ones you have to start. Also on FC6 and later you > can experiment with "virsh" for management of inactive domains. In particular you can do virsh list --all or virsh list --inactive And once you know the name, virsh start Or to delete a failed install virsh undefine This is available in FC6 or F7 (basically need libvirt 0.1.10 or later) Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From takatom at jp.fujitsu.com Fri Mar 9 01:20:14 2007 From: takatom at jp.fujitsu.com (Takahashi Tomohiro) Date: Fri, 9 Mar 2007 10:20:14 +0900 Subject: [et-mgmt-tools] [PATCH][RFC] Add --keymap option to virt-install Message-ID: <006301c761e9$1da62350$ecb1220a@fj98229700> Hi, Dan Thank you for committing a patch. I confirmed that I can use jp106 keyboard on HVM domain with virt-install and Error(KeyError: 'keymap') didn't happen with virt-manager. when I used latest patch(changeset 117). We also think it's automatically pick the correct keymap based on the X keymap. So, I intend to inquire into how to pick the correct keymap by Python. Please tell me how to pick the correct keymap by Python. if you know it. Thanks, Tomohiro Takahashi. > On Thu, Mar 08, 2007 at 09:14:19PM +0900, Takahashi Tomohiro wrote: >> Hi, >> >> Previously I tested keymap patch on virt-inst which was already posted. >> Today I test the patch on virt-manager.Then I found a bug. >> This patch fixes this problem. >> >> I would like to post an update patch. Because I caught a bug in keymap >> patch >> that I had posted on March 6. Errors did happen, when I used >> virt-manager. >> Because It doesn't have a function for keymap. > > Hmm, now that we have this virt-install patch, we'll need some way to let > people set the keymap from virt-manager too. I imagine we could add some > user preference to the 'Preferences' dialog box, either a plain text box > where they can enter a keymap, or perhaps a drop down list pre-populated > with all keymaps. Unless there is some way to automatically pick the > correct keymap based on the X keymap - but i don't think we get get at > that info any way. > >> virtmanager does not have keymap setting function. >> So I add a check "keymap" exists or not. > > Applied this updated patch to virt-install HG repo now.. > > http://hg.et.redhat.com/virt/applications/virtinst--devel?cs=660a6db5a4be > > Regards, > Dan. > -- > |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 > 2496 -=| > |=- Perl modules: > .cpan.org/~danberr/ -=| > |=- Projects: > at.net/~danielpb/ -=| > |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B > 505 -=| > From berrange at redhat.com Fri Mar 9 01:39:48 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Fri, 9 Mar 2007 01:39:48 +0000 Subject: [et-mgmt-tools] [PATCH][RFC] Add --keymap option to virt-install In-Reply-To: <006301c761e9$1da62350$ecb1220a@fj98229700> References: <006301c761e9$1da62350$ecb1220a@fj98229700> Message-ID: <20070309013948.GC10363@redhat.com> On Fri, Mar 09, 2007 at 10:20:14AM +0900, Takahashi Tomohiro wrote: > Hi, Dan > > Thank you for committing a patch. > > I confirmed that I can use jp106 keyboard on HVM domain with virt-install > and Error(KeyError: 'keymap') didn't happen with virt-manager. > when I used latest patch(changeset 117). > > We also think it's automatically pick the correct keymap based on the X > keymap. > So, I intend to inquire into how to pick the correct keymap by Python. > Please tell me how to pick the correct keymap by Python. if you know it. No, I've not found any way to automatically discover the keymap from Python / X / GTK yet... I'll let you know if I do. Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From ken at robson.net Sun Mar 11 11:14:47 2007 From: ken at robson.net (Ken Robson) Date: Sun, 11 Mar 2007 11:14:47 -0000 Subject: [et-mgmt-tools] Kernel Parameter Space Limits... In-Reply-To: <20070309013948.GC10363@redhat.com> References: <006301c761e9$1da62350$ecb1220a@fj98229700> <20070309013948.GC10363@redhat.com> Message-ID: <00b401c763ce$7da1ff90$0301a8c0@wxp1mbl1> Hi Folks, When evaluating Cobbler I noticed that it seems to be unaware of the maximum space limit for kernel parameters. Cobbler seems to generate a very large but highly meaningful parameters. This instantly caused me issues with my ks= directive being truncated and my install failing. My distributions file looks like: --- - arch: x86 breed: redhat initrd: /var/www/cobbler/ks_mirror/redhat.fedora.6.core/cdrom/images/pxeboot/initrd. img kernel: /var/www/cobbler/ks_mirror/redhat.fedora.6.core/cdrom/images/pxeboot/vmlinuz kernel_options: {} ks_meta: {} name: var_www_cobbler_ks_mirror_redhat.fedora.6.core_cdrom_images_pxeboot - arch: x86 breed: redhat initrd: /var/www/cobbler/ks_mirror/redhat.fedora.6.core/cdrom/images/xen/initrd.img kernel: /var/www/cobbler/ks_mirror/redhat.fedora.6.core/cdrom/images/xen/vmlinuz kernel_options: {} ks_meta: {} name: var_www_cobbler_ks_mirror_redhat.fedora.6.core_cdrom_images_xen The errant /tftpboot/pxelinux.cfg/default files looks like: default linux prompt 0 timeout 1 label linux kernel /images/var_www_cobbler_ks_mirror_redhat.fedora.6.core_cdrom_images_pxeboot/ vmlinuz append ksdevice=eth0 lang=en_US text syslog=192.168.1.133:25150 devfs=nomount ramdisk_size=16438 append initrd=/images/var_www_cobbler_ks_mirror_redhat.fedora.6.core_cdrom_images_p xeboot/initrd.img ks=http://192.168.1.133/cobbler_track/kickstarts_sys/default/ks.cfg And the settings file created by the import looks like: --- bootloaders: ia64: /var/lib/cobbler/elilo-3.6-ia64.efi standard: /usr/lib/syslinux/pxelinux.0 default_kickstart: /etc/cobbler/default.ks dhcpd_bin: /usr/sbin/dhcpd dhcpd_conf: /etc/dhcpd.conf httpd_bin: /usr/sbin/httpd kernel_options: append: ~ devfs: nomount ksdevice: eth0 lang: en_US ramdisk_size: 16438 text: ~ koan_path: /usr/bin/koan manage_dhcp: 1 minimize_syncs: 1 next_server: '192.168.1.133' server: '192.168.1.133' syslog_port: 25150 tftpboot: /tftpboot tftpd_bin: /usr/sbin/in.tftpd tftpd_conf: /etc/xinetd.d/tftp webdir: /var/www/cobbler When I try to kickstart using this set-up the append line get truncated and the install fails. If I alter the settings file to the following it works fine: --- bootloaders: ia64: /var/lib/cobbler/elilo-3.6-ia64.efi standard: /usr/lib/syslinux/pxelinux.0 default_kickstart: /etc/cobbler/default.ks dhcpd_bin: /usr/sbin/dhcpd dhcpd_conf: /etc/dhcpd.conf httpd_bin: /usr/sbin/httpd kernel_options: devfs: nomount ksdevice: eth0 lang: en_GB ramdisk_size: 16438 text: ~ koan_path: /usr/bin/koan manage_dhcp: 1 minimize_syncs: 1 next_server: '192.168.1.133' server: '192.168.1.133' syslog_port: 25150 tftpboot: /tftpboot tftpd_bin: /usr/sbin/in.tftpd tftpd_conf: /etc/xinetd.d/tftp webdir: /var/www/cobbler Thanks for your help. Kind regards, Ken. -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 3181 bytes Desc: not available URL: From ken at robson.net Sun Mar 11 11:24:14 2007 From: ken at robson.net (Ken Robson) Date: Sun, 11 Mar 2007 11:24:14 -0000 Subject: [et-mgmt-tools] Feature Request: Dynamic DHCP Updates Using OMAPI... Message-ID: <00c201c763cf$ccf029e0$0301a8c0@wxp1mbl1> Hi Folks, Rather than rewriting the dhcpd.conf file would it be possible for the DHCP management function to utilise OMAPI to interact with the DHCP server. In large Enterprises its just not possible to continually restart the DHCP server to add servers to the configuration. Thanks for your time and help. Kind regards, Ken. -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 3181 bytes Desc: not available URL: From ken at robson.net Sun Mar 11 16:37:20 2007 From: ken at robson.net (Ken Robson) Date: Sun, 11 Mar 2007 16:37:20 -0000 Subject: [et-mgmt-tools] Issues With $tree... Message-ID: <010601c763fb$8d5186e0$0301a8c0@wxp1mbl1> Hi Folks, When evaluating Cobbler I seem to be having problems with Cheetah substituting $tree in my /etc/cobbler/default.ks file. In my /etc/cobbler/default.ks file I see: # Use network installation url --url=$tree # If any cobbler repo definitions were referenced in the kickstart profile, include them here. $yum_repo_stanza In my /var/www/cobbler/kickstarts_sys/default/ks.cfg file I see: # Use network installation url --url=$tree # If any cobbler repo definitions were referenced in the kickstart profile, include them here. It appears that $yum_repo_stanza is being replaced (it is currently undefined in my set-up) but $tree is not resolving correctly. Thanks for any help and pointers. Kind regards, Ken. -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 3181 bytes Desc: not available URL: From ken at robson.net Sun Mar 11 17:59:01 2007 From: ken at robson.net (Ken Robson) Date: Sun, 11 Mar 2007 17:59:01 -0000 Subject: [et-mgmt-tools] Problems Attempting To Add A Profile With A Repository... Message-ID: <011d01c76406$f3b00910$0301a8c0@wxp1mbl1> Hi Folks, When trying to add a profile that includes a repository I seem to have the following issue: [root at lonvmmas01 cobbler]# cobbler profile add --name=redhat.fedora.6.i386.pxe --distro=var_www_cobbler_ks_mirror_redhat.fedora.6.core_cdrom_images_pxeboot --repos="redhat_fedora_6_updates" Traceback (most recent call last): File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 505, in main BootCLI(sys.argv).run() File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 102, in run self.curry_args(self.args[1:], self.commands['toplevel']) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 421, in curry_args commands[args[0]](args[1:]) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 475, in profile return self.curry_args(args, self.commands['profile']) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 421, in curry_args commands[args[0]](args[1:]) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 353, in profile_edit return self.apply_args(args,commands,on_ok) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 406, in apply_args input_routines[key](value) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 350, in '--repos' : lambda(a) : profile.set_repos(a) File "/usr/lib/python2.4/site-packages/cobbler/item_profile.py", line 93, in set_repos repolist.remove('') ValueError: list.remove(x): x not in list [root at lonvmmas01 cobbler]# However if I put a space following my repository name it then appears to work but seemingly then corrupts the repos key with an invalid value in the repos file: [root at lonvmmas01 cobbler]# cobbler profile add --name=redhat.fedora.6.i386.pxe --distro=var_www_cobbler_ks_mirror_redhat.fedora.6.core_cdrom_images_pxeboot --repos="redhat_fedora_6_updates " [root at lonvmmas01 cobbler]# cobbler sync one of the listed repositories is not defined in cobbler [root at lonvmmas01 cobbler]# Python is not my strong suit so any help would be appreciated. Kind regards, Ken. -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 3181 bytes Desc: not available URL: From fukuta.saori at jp.fujitsu.com Mon Mar 12 06:22:39 2007 From: fukuta.saori at jp.fujitsu.com (Saori Fukuta) Date: Mon, 12 Mar 2007 15:22:39 +0900 Subject: [et-mgmt-tools] [PATCH][RESEND] check of the required arguments In-Reply-To: <20070306180056.6F31.FUKUTA.SAORI@jp.fujitsu.com> References: <20070306180056.6F31.FUKUTA.SAORI@jp.fujitsu.com> Message-ID: <20070312152209.28EC.FUKUTA.SAORI@jp.fujitsu.com> Hi, I have no reply about following patch. So, I made it again for the current virt-install. I am looking forward to hearing your comment. If everything is OK, please apply it. On Tue, 06 Mar 2007 18:09:01 +0900 Saori Fukuta wrote: > So, here's what I'm going to suggest. I add two functions that check > the length of values, and if the value type is string I defined them > as a callback options . > They are called by following option: > long option (short option) > --name (-n) > --uuid (-u) > --file (-f) > --mac (-m) > --bridge (-b) > --connect > --cdrom (-c) > --os-type > --os-variant > --arch > --location (-l) add "--keymap" > > There is one exception. "--extra-args" doesn't call, because the option > has default value and that is ''. Thanks, Saori Fukuta. Index: virt-install =================================================================== diff -r 2e5b60ecbd93 virt-install --- a/virt-install Thu Mar 08 07:40:55 2007 -0500 +++ b/virt-install Mon Mar 12 14:45:30 2007 +0900 @@ -15,7 +15,7 @@ import os, sys, string -from optparse import OptionParser +from optparse import OptionParser, OptionValueError import subprocess import struct import logging @@ -220,20 +220,32 @@ def get_fullvirt_cdrom(cdpath, guest): cdpath = None ### Option parsing +def check_before_store(option, opt_str, value, parser): + if len(value) == 0: + raise OptionValueError, "%s option requires an argument" %opt_str + setattr(parser.values, option.dest, value) + +def check_before_append(option, opt_str, value, parser): + if len(value) == 0: + raise OptionValueError, "%s option requires an argument" %opt_str + parser.values.ensure_value(option.dest, []).append(value) + def parse_args(): parser = OptionParser() parser.add_option("-n", "--name", type="string", dest="name", + action="callback", callback=check_before_store, help="Name of the guest instance") parser.add_option("-r", "--ram", type="int", dest="memory", help="Memory to allocate for guest instance in megabytes") parser.add_option("-u", "--uuid", type="string", dest="uuid", + action="callback", callback=check_before_store, help="UUID for the guest; if none is given a random UUID will be generated") parser.add_option("", "--vcpus", type="int", dest="vcpus", help="Number of vcpus to configure for your guest") # disk options parser.add_option("-f", "--file", type="string", - dest="diskfile", action="append", + dest="diskfile", action="callback", callback=check_before_append, help="File to use as the disk image") parser.add_option("-s", "--file-size", type="float", action="append", dest="disksize", @@ -244,10 +256,10 @@ def parse_args(): # network options parser.add_option("-m", "--mac", type="string", - dest="mac", action="append", + dest="mac", action="callback", callback=check_before_append, help="Fixed MAC address for the guest; if none or RANDOM is given a random address will be used") parser.add_option("-b", "--bridge", type="string", - dest="bridge", action="append", + dest="bridge", action="callback", callback=check_before_append, help="Bridge to connect guest NIC to; if none given, will try to determine the default") # graphics options @@ -264,28 +276,38 @@ def parse_args(): help="Don't automatically try to connect to the guest console") parser.add_option("-k", "--keymap", type="string", dest="keymap", + action="callback", callback=check_before_store, help="set up keymap for a graphical console") parser.add_option("", "--accelerate", action="store_true", dest="accelerate", help="Use kernel acceleration capabilities") parser.add_option("", "--connect", type="string", dest="connect", + action="callback", callback=check_before_store, help="Connect to hypervisor with URI") # fullvirt options parser.add_option("-v", "--hvm", action="store_true", dest="fullvirt", help="This guest should be a fully virtualized guest") parser.add_option("-c", "--cdrom", type="string", dest="cdrom", + action="callback", callback=check_before_store, help="File to use a virtual CD-ROM device for fully virtualized guests") - parser.add_option("", "--os-type", type="string", dest="os_type", help="The OS type for fully virtualized guests, e.g. Linux, Solaris, Windows", default="Other") - parser.add_option("", "--os-variant", type="string", dest="os_variant", help="The OS variant for fully virtualized guests, e.g. Fedora, Solaris 8, Windows XP", default="Other") + parser.add_option("", "--os-type", type="string", dest="os_type", + action="callback", callback=check_before_store, + help="The OS type for fully virtualized guests, e.g. Linux, Solaris, Windows", default="Other") + parser.add_option("", "--os-variant", type="string", dest="os_variant", + action="callback", callback=check_before_store, + help="The OS variant for fully virtualized guests, e.g. Fedora, Solaris 8, Windows XP", default="Other") parser.add_option("", "--noapic", action="store_true", dest="noapic", help="Disables APIC for fully virtualized guest (overrides value in os-type/os-variant db)", default=False) parser.add_option("", "--noacpi", action="store_true", dest="noacpi", help="Disables ACPI for fully virtualized guest (overrides value in os-type/os-variant db)", default=False) - parser.add_option("", "--arch", type="string", dest="arch", help="The CPU architecture to simulate") + parser.add_option("", "--arch", type="string", dest="arch", + action="callback", callback=check_before_store, + help="The CPU architecture to simulate") # paravirt options parser.add_option("-p", "--paravirt", action="store_false", dest="fullvirt", help="This guest should be a paravirtualized guest") parser.add_option("-l", "--location", type="string", dest="location", + action="callback", callback=check_before_store, help="Installation source for paravirtualized guest (eg, nfs:host:/path, http://host/path, ftp://host/path)") parser.add_option("-x", "--extra-args", type="string", dest="extra", default="", From hbrock at redhat.com Mon Mar 12 13:40:20 2007 From: hbrock at redhat.com (Hugh Brock) Date: Mon, 12 Mar 2007 09:40:20 -0400 Subject: [et-mgmt-tools] [PATCH][RESEND] check of the required arguments In-Reply-To: <20070312152209.28EC.FUKUTA.SAORI@jp.fujitsu.com> References: <20070306180056.6F31.FUKUTA.SAORI@jp.fujitsu.com> <20070312152209.28EC.FUKUTA.SAORI@jp.fujitsu.com> Message-ID: <45F55844.3050802@redhat.com> Saori Fukuta wrote: > Hi, > > I have no reply about following patch. > So, I made it again for the current virt-install. > > I am looking forward to hearing your comment. > If everything is OK, please apply it. > > On Tue, 06 Mar 2007 18:09:01 +0900 Saori Fukuta wrote: > >> So, here's what I'm going to suggest. I add two functions that check >> the length of values, and if the value type is string I defined them >> as a callback options . >> They are called by following option: >> long option (short option) >> --name (-n) >> --uuid (-u) >> --file (-f) >> --mac (-m) >> --bridge (-b) >> --connect >> --cdrom (-c) >> --os-type >> --os-variant >> --arch >> --location (-l) > add "--keymap" > >> There is one exception. "--extra-args" doesn't call, because the option >> has default value and that is ''. > > Thanks, > Saori Fukuta. > > Index: virt-install > =================================================================== > diff -r 2e5b60ecbd93 virt-install > --- a/virt-install Thu Mar 08 07:40:55 2007 -0500 > +++ b/virt-install Mon Mar 12 14:45:30 2007 +0900 > @@ -15,7 +15,7 @@ > > > import os, sys, string > -from optparse import OptionParser > +from optparse import OptionParser, OptionValueError > import subprocess > import struct > import logging > @@ -220,20 +220,32 @@ def get_fullvirt_cdrom(cdpath, guest): > cdpath = None > > ### Option parsing > +def check_before_store(option, opt_str, value, parser): > + if len(value) == 0: > + raise OptionValueError, "%s option requires an argument" %opt_str > + setattr(parser.values, option.dest, value) > + > +def check_before_append(option, opt_str, value, parser): > + if len(value) == 0: > + raise OptionValueError, "%s option requires an argument" %opt_str > + parser.values.ensure_value(option.dest, []).append(value) > + > def parse_args(): > parser = OptionParser() > parser.add_option("-n", "--name", type="string", dest="name", > + action="callback", callback=check_before_store, > help="Name of the guest instance") > parser.add_option("-r", "--ram", type="int", dest="memory", > help="Memory to allocate for guest instance in megabytes") > parser.add_option("-u", "--uuid", type="string", dest="uuid", > + action="callback", callback=check_before_store, > help="UUID for the guest; if none is given a random UUID will be generated") > parser.add_option("", "--vcpus", type="int", dest="vcpus", > help="Number of vcpus to configure for your guest") > > # disk options > parser.add_option("-f", "--file", type="string", > - dest="diskfile", action="append", > + dest="diskfile", action="callback", callback=check_before_append, > help="File to use as the disk image") > parser.add_option("-s", "--file-size", type="float", > action="append", dest="disksize", > @@ -244,10 +256,10 @@ def parse_args(): > > # network options > parser.add_option("-m", "--mac", type="string", > - dest="mac", action="append", > + dest="mac", action="callback", callback=check_before_append, > help="Fixed MAC address for the guest; if none or RANDOM is given a random address will be used") > parser.add_option("-b", "--bridge", type="string", > - dest="bridge", action="append", > + dest="bridge", action="callback", callback=check_before_append, > help="Bridge to connect guest NIC to; if none given, will try to determine the default") > > # graphics options > @@ -264,28 +276,38 @@ def parse_args(): > help="Don't automatically try to connect to the guest console") > > parser.add_option("-k", "--keymap", type="string", dest="keymap", > + action="callback", callback=check_before_store, > help="set up keymap for a graphical console") > > parser.add_option("", "--accelerate", action="store_true", dest="accelerate", > help="Use kernel acceleration capabilities") > parser.add_option("", "--connect", type="string", dest="connect", > + action="callback", callback=check_before_store, > help="Connect to hypervisor with URI") > > # fullvirt options > parser.add_option("-v", "--hvm", action="store_true", dest="fullvirt", > help="This guest should be a fully virtualized guest") > parser.add_option("-c", "--cdrom", type="string", dest="cdrom", > + action="callback", callback=check_before_store, > help="File to use a virtual CD-ROM device for fully virtualized guests") > - parser.add_option("", "--os-type", type="string", dest="os_type", help="The OS type for fully virtualized guests, e.g. Linux, Solaris, Windows", default="Other") > - parser.add_option("", "--os-variant", type="string", dest="os_variant", help="The OS variant for fully virtualized guests, e.g. Fedora, Solaris 8, Windows XP", default="Other") > + parser.add_option("", "--os-type", type="string", dest="os_type", > + action="callback", callback=check_before_store, > + help="The OS type for fully virtualized guests, e.g. Linux, Solaris, Windows", default="Other") > + parser.add_option("", "--os-variant", type="string", dest="os_variant", > + action="callback", callback=check_before_store, > + help="The OS variant for fully virtualized guests, e.g. Fedora, Solaris 8, Windows XP", default="Other") > parser.add_option("", "--noapic", action="store_true", dest="noapic", help="Disables APIC for fully virtualized guest (overrides value in os-type/os-variant db)", default=False) > parser.add_option("", "--noacpi", action="store_true", dest="noacpi", help="Disables ACPI for fully virtualized guest (overrides value in os-type/os-variant db)", default=False) > - parser.add_option("", "--arch", type="string", dest="arch", help="The CPU architecture to simulate") > + parser.add_option("", "--arch", type="string", dest="arch", > + action="callback", callback=check_before_store, > + help="The CPU architecture to simulate") > > # paravirt options > parser.add_option("-p", "--paravirt", action="store_false", dest="fullvirt", > help="This guest should be a paravirtualized guest") > parser.add_option("-l", "--location", type="string", dest="location", > + action="callback", callback=check_before_store, > help="Installation source for paravirtualized guest (eg, nfs:host:/path, http://host/path, ftp://host/path)") > parser.add_option("-x", "--extra-args", type="string", > dest="extra", default="", > > Hi! Sorry, I thought we had already applied this. I'm testing it now and will apply it if no problems. Thanks, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From mdehaan at redhat.com Mon Mar 12 13:54:45 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 12 Mar 2007 09:54:45 -0400 Subject: [et-mgmt-tools] Kernel Parameter Space Limits... In-Reply-To: <00b401c763ce$7da1ff90$0301a8c0@wxp1mbl1> References: <006301c761e9$1da62350$ecb1220a@fj98229700> <20070309013948.GC10363@redhat.com> <00b401c763ce$7da1ff90$0301a8c0@wxp1mbl1> Message-ID: <45F55BA5.2040505@redhat.com> Ken Robson wrote: > Hi Folks, > > When evaluating Cobbler I noticed that it seems to be unaware of the maximum > space limit for kernel parameters. Do you know what the actual limit is? I'll admit I don't, and I haven't seen this before... Could this be syslinux causing trouble and not the actual kernel? The "append" stuff can certainly be removed. > Cobbler seems to generate a very large > but highly meaningful parameters. This instantly caused me issues with my > ks= directive being truncated and my install failing. My distributions file > looks like: > > --- > - > arch: x86 > breed: redhat > initrd: > /var/www/cobbler/ks_mirror/redhat.fedora.6.core/cdrom/images/pxeboot/initrd. > img > kernel: > /var/www/cobbler/ks_mirror/redhat.fedora.6.core/cdrom/images/pxeboot/vmlinuz > kernel_options: {} > ks_meta: {} > name: > var_www_cobbler_ks_mirror_redhat.fedora.6.core_cdrom_images_pxeboot > - > arch: x86 > breed: redhat > initrd: > /var/www/cobbler/ks_mirror/redhat.fedora.6.core/cdrom/images/xen/initrd.img > kernel: > /var/www/cobbler/ks_mirror/redhat.fedora.6.core/cdrom/images/xen/vmlinuz > kernel_options: {} > ks_meta: {} > name: var_www_cobbler_ks_mirror_redhat.fedora.6.core_cdrom_images_xen > > The errant /tftpboot/pxelinux.cfg/default files looks like: > > default linux > prompt 0 > timeout 1 > label linux > kernel > /images/var_www_cobbler_ks_mirror_redhat.fedora.6.core_cdrom_images_pxeboot/ > vmlinuz > append ksdevice=eth0 lang=en_US text syslog=192.168.1.133:25150 > devfs=nomount ramdisk_size=16438 append > > initrd=/images/var_www_cobbler_ks_mirror_redhat.fedora.6.core_cdrom_images_p > xeboot/initrd.img > ks=http://192.168.1.133/cobbler_track/kickstarts_sys/default/ks.cfg > > And the settings file created by the import looks like: > > --- > bootloaders: > ia64: /var/lib/cobbler/elilo-3.6-ia64.efi > standard: /usr/lib/syslinux/pxelinux.0 > default_kickstart: /etc/cobbler/default.ks > dhcpd_bin: /usr/sbin/dhcpd > dhcpd_conf: /etc/dhcpd.conf > httpd_bin: /usr/sbin/httpd > kernel_options: > append: ~ > devfs: nomount > ksdevice: eth0 > lang: en_US > ramdisk_size: 16438 > text: ~ > koan_path: /usr/bin/koan > manage_dhcp: 1 > minimize_syncs: 1 > next_server: '192.168.1.133' > server: '192.168.1.133' > syslog_port: 25150 > tftpboot: /tftpboot > tftpd_bin: /usr/sbin/in.tftpd > tftpd_conf: /etc/xinetd.d/tftp > webdir: /var/www/cobbler > > When I try to kickstart using this set-up the append line get truncated and > the install fails. If I alter the settings file to the following it works > fine: > > --- > bootloaders: > ia64: /var/lib/cobbler/elilo-3.6-ia64.efi > standard: /usr/lib/syslinux/pxelinux.0 > default_kickstart: /etc/cobbler/default.ks > dhcpd_bin: /usr/sbin/dhcpd > dhcpd_conf: /etc/dhcpd.conf > httpd_bin: /usr/sbin/httpd > kernel_options: > devfs: nomount > ksdevice: eth0 > lang: en_GB > ramdisk_size: 16438 > text: ~ > koan_path: /usr/bin/koan > manage_dhcp: 1 > minimize_syncs: 1 > next_server: '192.168.1.133' > server: '192.168.1.133' > syslog_port: 25150 > tftpboot: /tftpboot > tftpd_bin: /usr/sbin/in.tftpd > tftpd_conf: /etc/xinetd.d/tftp > webdir: /var/www/cobbler > > Thanks for your help. > > Kind regards, > > Ken. > > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From mdehaan at redhat.com Mon Mar 12 14:03:33 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 12 Mar 2007 10:03:33 -0400 Subject: [et-mgmt-tools] Feature Request: Dynamic DHCP Updates Using OMAPI... In-Reply-To: <00c201c763cf$ccf029e0$0301a8c0@wxp1mbl1> References: <00c201c763cf$ccf029e0$0301a8c0@wxp1mbl1> Message-ID: <45F55DB5.90705@redhat.com> Ken Robson wrote: > Hi Folks, > > Rather than rewriting the dhcpd.conf file would it be possible for the DHCP > management function to utilise OMAPI to interact with the DHCP server. In > large Enterprises its just not possible to continually restart the DHCP > server to add servers to the configuration. > > Thanks for your time and help. > > Kind regards, > > Ken. > > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > > https://www.redhat.com/mailman/listinfo/et-mgmt-tools This looks like it would be pretty interesting. I'll put it on the "investigate" list, though if you'd like to take a crack at adding support for it, be my guest. --Michael From mdehaan at redhat.com Mon Mar 12 14:27:23 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 12 Mar 2007 10:27:23 -0400 Subject: [et-mgmt-tools] Issues With $tree... In-Reply-To: <010601c763fb$8d5186e0$0301a8c0@wxp1mbl1> References: <010601c763fb$8d5186e0$0301a8c0@wxp1mbl1> Message-ID: <45F5634B.9020909@redhat.com> Ken Robson wrote: > Hi Folks, > > When evaluating Cobbler I seem to be having problems with Cheetah > substituting $tree in my /etc/cobbler/default.ks file. In my > /etc/cobbler/default.ks file I see: > > # Use network installation > url --url=$tree > # If any cobbler repo definitions were referenced in the kickstart > profile, include > them here. > $yum_repo_stanza > > In my /var/www/cobbler/kickstarts_sys/default/ks.cfg file I see: > > # Use network installation > url --url=$tree > # If any cobbler repo definitions were referenced in the kickstart > profile, include > them here. > > It appears that $yum_repo_stanza is being replaced (it is currently > undefined in my set-up) but $tree is not resolving correctly. > > Thanks for any help and pointers. > > Kind regards, > > Ken. > Thanks for the feedback... It looks like you were using "cobbler import --mirror" on a DVD location, or perhaps the older "cobbler import --path". Which was it? If it was "--mirror", that code, being designed for public rsync mirrors and not DVD's, likely did not have enough information to pick a good kickstart for the distro you used. If it was "--path", that code doesn't auto-assign kickstarts at all. Basically cobbler uses known pieces of the path, which are found on all public mirrors, like "FC-6" or "RHEL-4" to determine what the distro is and to assign a good kickstart to that distro. Either way, you can edit the profile file (/var/lib/cobbler/profiles) to add in a tree location to use. It's another element of "ksmeta" and just needs to have a value like "http://server/cobbler_track/path/to/files/os" or equivalent. I'll go back and see if the kickstart assignment code can be tweaked to better suit the DVD import cases. > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From Curtis at GreenKey.net Mon Mar 12 17:30:30 2007 From: Curtis at GreenKey.net (Curtis Doty) Date: Mon, 12 Mar 2007 10:30:30 -0700 (PDT) Subject: [et-mgmt-tools] Kernel Parameter Space Limits... In-Reply-To: <45F55BA5.2040505@redhat.com> References: <006301c761e9$1da62350$ecb1220a@fj98229700> <20070309013948.GC10363@redhat.com> <00b401c763ce$7da1ff90$0301a8c0@wxp1mbl1> <45F55BA5.2040505@redhat.com> Message-ID: <20070312173030.EC9CD6F06C@alopias.GreenKey.net> 9:54am Michael DeHaan said: > Ken Robson wrote: > > Hi Folks, > > > > When evaluating Cobbler I noticed that it seems to be unaware of the maximum > > space limit for kernel parameters. > > Do you know what the actual limit is? I'll admit I don't, and I haven't seen > this before... Could this be syslinux causing trouble and not the actual > kernel? According to syslinux, the kernel limit is 255. http://google.com/search?q=syslinux+max_cmd_len ../C From hbrock at redhat.com Mon Mar 12 21:31:40 2007 From: hbrock at redhat.com (Hugh Brock) Date: Mon, 12 Mar 2007 17:31:40 -0400 Subject: [et-mgmt-tools] [PATCH][RESEND] check of the required arguments In-Reply-To: <20070312152209.28EC.FUKUTA.SAORI@jp.fujitsu.com> References: <20070306180056.6F31.FUKUTA.SAORI@jp.fujitsu.com> <20070312152209.28EC.FUKUTA.SAORI@jp.fujitsu.com> Message-ID: <45F5C6BC.70602@redhat.com> Saori Fukuta wrote: > Hi, > > I have no reply about following patch. > So, I made it again for the current virt-install. > > I am looking forward to hearing your comment. > If everything is OK, please apply it. > > On Tue, 06 Mar 2007 18:09:01 +0900 Saori Fukuta wrote: > >> So, here's what I'm going to suggest. I add two functions that check >> the length of values, and if the value type is string I defined them >> as a callback options . >> They are called by following option: >> long option (short option) >> --name (-n) >> --uuid (-u) >> --file (-f) >> --mac (-m) >> --bridge (-b) >> --connect >> --cdrom (-c) >> --os-type >> --os-variant >> --arch >> --location (-l) > add "--keymap" > >> There is one exception. "--extra-args" doesn't call, because the option >> has default value and that is ''. > Thanks Saori-san, patch applied. Appears to work perfectly. --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From fukuta.saori at jp.fujitsu.com Tue Mar 13 00:05:40 2007 From: fukuta.saori at jp.fujitsu.com (Saori Fukuta) Date: Tue, 13 Mar 2007 09:05:40 +0900 Subject: [et-mgmt-tools] [PATCH][RESEND] check of the required arguments In-Reply-To: <45F5C6BC.70602@redhat.com> References: <20070312152209.28EC.FUKUTA.SAORI@jp.fujitsu.com> <45F5C6BC.70602@redhat.com> Message-ID: <20070313084729.9D83.FUKUTA.SAORI@jp.fujitsu.com> On Mon, 12 Mar 2007 17:31:40 -0400 Hugh Brock wrote: > Saori Fukuta wrote: > > Hi, > > > > I have no reply about following patch. > > So, I made it again for the current virt-install. > > > > I am looking forward to hearing your comment. > > If everything is OK, please apply it. > > > > On Tue, 06 Mar 2007 18:09:01 +0900 Saori Fukuta wrote: > > > >> So, here's what I'm going to suggest. I add two functions that check > >> the length of values, and if the value type is string I defined them > >> as a callback options . > >> They are called by following option: > >> long option (short option) > >> --name (-n) > >> --uuid (-u) > >> --file (-f) > >> --mac (-m) > >> --bridge (-b) > >> --connect > >> --cdrom (-c) > >> --os-type > >> --os-variant > >> --arch > >> --location (-l) > > add "--keymap" > > > >> There is one exception. "--extra-args" doesn't call, because the option > >> has default value and that is ''. > > > > Thanks Saori-san, patch applied. Appears to work perfectly. > --Hugh Thank you Hugh. >From now on, we can use this callback when value type of option argument is string. :-) Saori. From fj0873gn at aa.jp.fujitsu.com Tue Mar 13 07:08:25 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Tue, 13 Mar 2007 16:08:25 +0900 Subject: [et-mgmt-tools] [PATCH][RESEND] "--file-size" option is ignored In-Reply-To: <200703061230.DHG04272.0G849HOK@aa.jp.fujitsu.com> References: <200703061230.DHG04272.0G849HOK@aa.jp.fujitsu.com> Message-ID: <200703131608.AAC82321.8G40OK9H@aa.jp.fujitsu.com> Hi, Dan Would you give me a comment on this patch? If not, please apply it. > When I execute "virt-install --file-size=10", > the value of "--file-size" option is ignored and the error message > "ERROR: float() argument must be a string or a number" is output > after I answer "What would you like to use as the disk (path)?". > > The attached patch adds to solve this problem. > > > Signed-off-by: Nobuhiro Itou Thanks, Nobuhiro Itou. diff -r 95196b0f37a5 virt-install --- a/virt-install Mon Mar 12 12:39:59 2007 -0400 +++ b/virt-install Tue Mar 13 15:55:13 2007 +0900 @@ -106,7 +106,10 @@ def get_disk(disk, size, sparse, guest, def get_disk(disk, size, sparse, guest, hvm): # FIXME: need to handle a list of disks at some point while 1: - disk = prompt_for_input("What would you like to use as the disk (path)?", disk) + msg = "What would you like to use as the disk (path)?" + if not size is None: + msg = "What would you like the size %sGB to use as the disk (path)?" %(size,) + disk = prompt_for_input(msg, disk) while 1: if os.path.exists(disk): break @@ -138,8 +141,13 @@ def get_disks(disk, size, sparse, guest, sys.exit(1) elif type(disk) == list: size = [ None ] * len(disk) + elif type(size) == list: + disk = [ None ] * len(size) if (type(disk) == list): map(lambda d, s: get_disk(d, s, sparse, guest, hvm), disk, size) + elif (type(size) == list): + map(lambda d, s: get_disk(d, s, sparse, guest, hvm), + disk, size) else: From fj0873gn at aa.jp.fujitsu.com Tue Mar 13 07:16:56 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Tue, 13 Mar 2007 16:16:56 +0900 Subject: [et-mgmt-tools] [PATCH][RESEND] connecting console after install completes In-Reply-To: <200703071836.BHE35429.K9O840GH@aa.jp.fujitsu.com> References: <200703071836.BHE35429.K9O840GH@aa.jp.fujitsu.com> Message-ID: <200703131616.JGI57822.K09HG8O4@aa.jp.fujitsu.com> Hi, Dan Would you give me a comment on this patch? If not, please apply it. > When the domain restarts after install completes, the console is not connected. > Therefore, if I don't connect the console for myself, > I cannot be finished the setting after install completes. > > The attached patch adds to solve this problem. > > > Signed-off-by: Nobuhiro Itou Thanks, Nobuhiro Itou. diff -r 95196b0f37a5 virt-install --- a/virt-install Mon Mar 12 12:39:59 2007 -0400 +++ b/virt-install Tue Mar 13 16:09:12 2007 +0900 @@ -509,6 +509,7 @@ def main(): # the domain print "Guest installation complete... restarting guest." dom.create() + guest.connect_console(conscb) else: print ("Domain installation does not appear to have been\n" "successful. If it was, you can restart your domain\n" diff -r 95196b0f37a5 virtinst/Guest.py --- a/virtinst/Guest.py Mon Mar 12 12:39:59 2007 -0400 +++ b/virtinst/Guest.py Tue Mar 13 16:11:56 2007 +0900 @@ -575,6 +575,18 @@ class Guest(object): # for inactive guest, or get the still running install.. return self.conn.lookupByName(self.name) + def connect_console(self, consolecb): + child = None + if consolecb: + logging.debug("Launching console callback") + child = consolecb(self.domain) + + if child: # if we connected the console, wait for it to finish + try: + (pid, status) = os.waitpid(child, 0) + except OSError, (errno, msg): + print __name__, "waitpid:", msg + def validate_parms(self): if self.domain is not None: raise RuntimeError, "Domain already started!" From fj7716hz at aa.jp.fujitsu.com Tue Mar 13 08:11:06 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Tue, 13 Mar 2007 17:11:06 +0900 Subject: [et-mgmt-tools] [PATCH][RESEND}Check the starting domain's uuid In-Reply-To: <45E68AB0.9070908@aa.jp.fujitsu.com> References: <45E68AB0.9070908@aa.jp.fujitsu.com> Message-ID: <45F65C9A.1050806@aa.jp.fujitsu.com> Hi, Would you give me a comment on this patch? If not, please apply it. Tatsuro Enokura wrote: > The virt-install command has a problem that the information of > the existent domain is overwritten by virt-install command > specifing the domain's uuid. > Futhermore, the existent domain's set-up file is overwritten > on the xen-3.0.4. > > I make a patch checking the starting domain's uuid on Guest.py. Signed-off-by: Tatsuro Enokura Thanks, Tatsuro Enokura ------- diff -r 95196b0f37a5 virtinst/Guest.py --- a/virtinst/Guest.py Mon Mar 12 12:39:59 2007 -0400 +++ b/virtinst/Guest.py Tue Mar 13 17:11:14 2007 +0900 @@ -582,7 +582,19 @@ class Guest(object): def _set_defaults(self): if self.uuid is None: - self.uuid = util.uuidToString(util.randomUUID()) + while 1: + self.uuid = util.uuidToString(util.randomUUID()) + try: + if self.conn.lookupByUUIDString(self.uuid) is not None: + continue + except libvirt.libvirtError: + break + else: + try: + if self.conn.lookupByUUIDString(self.uuid) is not None: + raise RuntimeError, "UUID has been already used by the other guests!" + except libvirt.libvirtError: + pass if self.vcpus is None: self.vcpus = 1 if self.name is None or self.memory is None: ------------------------------------------------------ -- ==================================================================== Tatsuro Enokura Toyama Fujitsu Limited Software Development)2nd Development)2nd Project E-mail:fj7716hz at aa.jp.fujitsu.com Phone:(076)455-3131 Fax:(076)455-0956 Toll:7576-3622 ==================================================================== From fj7716hz at aa.jp.fujitsu.com Tue Mar 13 08:27:14 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Tue, 13 Mar 2007 17:27:14 +0900 Subject: [et-mgmt-tools] [PATCH] Check the making domain's mac address In-Reply-To: <45EE89B6.5010200@aa.jp.fujitsu.com> References: <45EE89B6.5010200@aa.jp.fujitsu.com> Message-ID: <45F66062.8010900@aa.jp.fujitsu.com> Hi, Would you give me a comment on this patch? If not, please apply it. Tatsuro Enokura wrote: > The virt-install command can specify the making domain's vnif > MAC address. The MAC address must be unique on the system, > but the virt-install command doesn't check that > the MAC address is unique among the running domains and host. > > The attached patch resolve this issue in the following way: > > 1) Get the running Domain's vnif MAC address. > 2) Get the host's NIC MAC address. > 3) Check the making domain's MAC address with 1) and 2) 's data. Signed-off-by: Tatsuro Enokura Thanks, Tatsuro Enokura --------------------------------------------------------------------- diff -r 95196b0f37a5 virtinst/Guest.py --- a/virtinst/Guest.py Mon Mar 12 12:39:59 2007 -0400 +++ b/virtinst/Guest.py Tue Mar 13 17:21:11 2007 +0900 @@ -148,9 +148,40 @@ class VirtualNetworkInterface: self.macaddr = macaddr self.bridge = bridge - def setup(self): + def setup(self, conn): + # get the running domain's vNIC MACaddresss + macaddrlist = [] + ids = conn.listDomainsID(); + for id in ids: + vm = conn.lookupByID(id) + disks = util.get_network_devices(vm) + for (dummy, dummy, dummy, vm_macaddr) in disks: + macaddrlist.append(vm_macaddr) + + # get the Host's NIC MACaddress + host_macaddrlist = [] + hostdevs = util.get_host_network_devices() + for(dummy, dummy, dummy, dummy, host_macaddr) in hostdevs: + host_macaddrlist.append(host_macaddr) + if self.macaddr is None: - self.macaddr = util.randomMAC() + # macaddrs used check + retryFlg = True + while retryFlg: + self.macaddr = util.randomMAC() + for macaddr in macaddrlist: + if self.macaddr.upper() == macaddr.upper(): + break + else: + retryFlg = False + else: + for macaddr in macaddrlist: + if self.macaddr.upper() == macaddr.upper(): + raise ValueError, "MAC address has been already used by the running domain." + for macaddr in host_macaddrlist: + if self.macaddr.upper() == macaddr.upper(): + raise ValueError, "MAC address has been collided with the physical NIC." + if not self.bridge: self.bridge = util.default_bridge() @@ -429,7 +460,7 @@ class Guest(object): for disk in self.disks: disk.setup(progresscb) for nic in self.nics: - nic.setup() + nic.setup(self.conn) def _get_disk_xml(self, install = True): """Get the disk config in the libvirt XML format""" diff -r 95196b0f37a5 virtinst/util.py --- a/virtinst/util.py Mon Mar 12 12:39:59 2007 -0400 +++ b/virtinst/util.py Tue Mar 13 17:18:23 2007 +0900 @@ -14,6 +14,7 @@ import random import os.path +import libxml2 from sys import stderr def default_bridge(): @@ -123,3 +124,60 @@ def uuidFromString(s): def uuidFromString(s): s = s.replace('-', '') return [ int(s[i : i + 2], 16) for i in range(0, 32, 2) ] + +# the following function is from virt-manager/src/virtManager/domain.py +def get_network_devices(vm): + xml = vm.XMLDesc(0) + doc = None + try: + doc = libxml2.parseDoc(xml) + except: + return [] + ctx = doc.xpathNewContext() + disks = [] + try: + ret = ctx.xpathEval("/domain/devices/interface") + + for node in ret: + type = node.prop("type") + devmac = None + source = None + for child in node.children: + if child.name == "source": + if type == "bridge": + source = child.prop("bridge") + elif child.name == "mac": + devmac = child.prop("address") + + if source == None: + source = "-" + + devdst = "eth%d" % len(disks) + + disks.append([type, source, devdst, devmac]) + finally: + if ctx != None: + ctx.xpathFreeContext() + if doc != None: + doc.freeDoc() + return disks + +# the following function quotes from python2.5/uuid.py +def get_host_network_devices(): + device = [] + for dir in ['', '/sbin/', '/usr/sbin']: + executable = os.path.join(dir, "ifconfig") + if not os.path.exists(executable): + continue + try: + cmd = 'LC_ALL=C %s -a 2>/dev/null' % (executable) + pipe = os.popen(cmd) + except IOError: + continue + for line in pipe: + words = line.lower().split() + for i in range(len(words)): + if words[i] == "hwaddr": + device.append(words) + return device + -------------------------------------------------------------------- From hbrock at redhat.com Tue Mar 13 13:09:30 2007 From: hbrock at redhat.com (Hugh Brock) Date: Tue, 13 Mar 2007 09:09:30 -0400 Subject: [et-mgmt-tools] [PATCH][RESEND] "--file-size" option is ignored In-Reply-To: <200703131608.AAC82321.8G40OK9H@aa.jp.fujitsu.com> References: <200703061230.DHG04272.0G849HOK@aa.jp.fujitsu.com> <200703131608.AAC82321.8G40OK9H@aa.jp.fujitsu.com> Message-ID: <45F6A28A.1090806@redhat.com> Nobuhiro Itou wrote: > Hi, Dan > > Would you give me a comment on this patch? > If not, please apply it. > >> When I execute "virt-install --file-size=10", >> the value of "--file-size" option is ignored and the error message >> "ERROR: float() argument must be a string or a number" is output >> after I answer "What would you like to use as the disk (path)?". >> >> The attached patch adds to solve this problem. >> >> >> Signed-off-by: Nobuhiro Itou > > > Thanks, > Nobuhiro Itou. > This looks fine, I'll go ahead and apply it. Thanks! --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From hbrock at redhat.com Tue Mar 13 13:14:16 2007 From: hbrock at redhat.com (Hugh Brock) Date: Tue, 13 Mar 2007 09:14:16 -0400 Subject: [et-mgmt-tools] [PATCH][RESEND] connecting console after install completes In-Reply-To: <200703131616.JGI57822.K09HG8O4@aa.jp.fujitsu.com> References: <200703071836.BHE35429.K9O840GH@aa.jp.fujitsu.com> <200703131616.JGI57822.K09HG8O4@aa.jp.fujitsu.com> Message-ID: <45F6A3A8.7040708@redhat.com> Nobuhiro Itou wrote: > Hi, Dan > > Would you give me a comment on this patch? > If not, please apply it. > >> When the domain restarts after install completes, the console is not connected. >> Therefore, if I don't connect the console for myself, >> I cannot be finished the setting after install completes. >> >> The attached patch adds to solve this problem. >> >> >> Signed-off-by: Nobuhiro Itou > > > Thanks, > Nobuhiro Itou. > Two issues here: 1. The whole "restart the guest automatically" feature in virt-install is a bit of a hack, and we are contemplating removing it altogether. The problem is that virt-install has no real way of knowing whether the guest that just shut down is really waiting to be rebooted, or is in fact a failed install that the user would not want restarted anyway. Having said this, as long as the "restart" hack is still there, I guess I have no problem reconnecting the console as well. 2. Minor nit: instead of swallowing the error like + except OSError, (errno, msg): + print __name__, "waitpid:", msg why not go ahead and raise a RuntimeError (and log the message) or something similar? If you want to fix 2. and resubmit, I'll go ahead and apply it. Thanks! --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From hbrock at redhat.com Tue Mar 13 14:14:05 2007 From: hbrock at redhat.com (Hugh Brock) Date: Tue, 13 Mar 2007 10:14:05 -0400 Subject: [et-mgmt-tools] [PATCH][RESEND}Check the starting domain's uuid In-Reply-To: <45F65C9A.1050806@aa.jp.fujitsu.com> References: <45E68AB0.9070908@aa.jp.fujitsu.com> <45F65C9A.1050806@aa.jp.fujitsu.com> Message-ID: <45F6B1AD.30005@redhat.com> Tatsuro Enokura wrote: > Hi, > > Would you give me a comment on this patch? > If not, please apply it. > > Tatsuro Enokura wrote: >> The virt-install command has a problem that the information of >> the existent domain is overwritten by virt-install command >> specifing the domain's uuid. >> Futhermore, the existent domain's set-up file is overwritten >> on the xen-3.0.4. >> >> I make a patch checking the starting domain's uuid on Guest.py. > > Signed-off-by: Tatsuro Enokura > > Thanks, > Tatsuro Enokura > This should be fine. The only problem I have with it is that the libvirt Python API is a bit broken on this point -- it should really return None without an error condition if the domain does not exist, rather than raising an error. In anticipation that the libvirt Python API will eventually be fixed to work properly in this case, I've added "else" clauses that handle the case where the lookup returns None but does not throw an error. With that minor change, applied. Thanks! --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From rnijskens at gmail.com Tue Mar 13 14:38:24 2007 From: rnijskens at gmail.com (rnijskens at gmail.com) Date: Tue, 13 Mar 2007 15:38:24 +0100 Subject: [et-mgmt-tools] libvirtError: virDomainCreateLinux() failed POST operation failed:, (xend.err "Error creating domain: (22, 'Invalid argument')") Message-ID: <45F6B760.7030602@gmail.com> Hi, Whatever I do, I get always this error: libvirtError: virDomainCreateLinux() failed POST operation failed: (xend.err "Error creating domain: (22, 'Invalid argument')") [root at matisse ~]# rpm -q cobbler cobbler-0.4.3-0.fc6 [root at matisse ~]# rpm -q koan koan-0.2.6-1.fc6 [root at matisse ~]# /etc/init.d/xend status xend is running [root at matisse ~]# xm list Name ID Mem(MiB) VCPUs State Time(s) Domain-0 0 1500 2 r----- 439.3 [root at matisse ~]# cobbler report distro : centos4.4_server kernel : /home/Centos/images/pxeboot/vmlinuz initrd : /home/Centos/images/pxeboot/initrd.img kernel options : {} architecture : x86 ks metadata : {} breed : redhat profile : webserver distro : centos4.4_server kickstart : /etc/cobbler/default.ks kernel options : {} ks metadata : {} virt name : virtguest virt file size : 5 virt ram : 512 virt paravirt : True repos : [root at matisse ~]# koan --virt --profile=webserver --server=localhost - processing profile: webserver - fetching configuration for profile: webserver - url=http://localhost/cobbler/profiles/webserver - {'kickstart': 'http://127.0.0.1/cobbler_track/kickstarts/webserver/ks.cfg', 'name': 'webserver', 'virt_ram': 512, 'repos': '', 'kernel_options': 'ksdevice=eth0 text syslog=127.0.0.1:25150 lang== devfs=nomount ramdisk_size=16438 append ', 'virt_name': 'virtguest', 'virt_file_size': 5, 'distro': 'centos4.4_server', 'virt_paravirt': 'True', 'ks_meta': ''} - fetching configuration for distro: centos4.4_server - url=http://localhost/cobbler/distros/centos4.4_server - {'kernel': '/home/Centos/images/pxeboot/vmlinuz', 'ks_meta': '', 'breed': 'redhat', 'kernel_options': 'ksdevice=eth0 text syslog=127.0.0.1:25150 lang== devfs=nomount ramdisk_size=16438 append ', 'initrd': '/home/Centos/images/pxeboot/initrd.img', 'arch': 'x86', 'name': 'centos4.4_server'} - downloading initrd initrd.img to /tmp/initrd.img - url=http://localhost/cobbler/images/centos4.4_server/initrd.img - downloading kernel vmlinuz to /tmp/vmlinuz - url=http://localhost/cobbler/images/centos4.4_server/vmlinuz - kernel saved = /tmp/vmlinuz - initrd saved = /tmp/initrd.img libvir: Xen Daemon error : POST operation failed: (xend.err "Error creating domain: (22, 'Invalid argument')") Failed to create domain virtguest Traceback (most recent call last): File "/usr/lib/python2.4/site-packages/koan/app.py", line 97, in main k.run() File "/usr/lib/python2.4/site-packages/koan/app.py", line 155, in run self.do_virt() File "/usr/lib/python2.4/site-packages/koan/app.py", line 300, in do_virt return self.do_net_install("/tmp",after_download) File "/usr/lib/python2.4/site-packages/koan/app.py", line 256, in do_net_install after_download(self, distro_data, profile_data) File "/usr/lib/python2.4/site-packages/koan/app.py", line 299, in after_download self.do_virt_net_install(profile_data, distro_data) File "/usr/lib/python2.4/site-packages/koan/app.py", line 599, in do_virt_net_install extra=kextra File "/usr/lib/python2.4/site-packages/koan/virtcreate.py", line 201, in start_paravirt_install dom = conn.createLinux(cfgxml, 0) File "/usr/lib/python2.4/site-packages/libvirt.py", line 329, in createLinux if ret is None:raise libvirtError('virDomainCreateLinux() failed', conn=self) libvirtError: virDomainCreateLinux() failed POST operation failed: (xend.err "Error creating domain: (22, 'Invalid argument')") From hbrock at redhat.com Tue Mar 13 14:45:17 2007 From: hbrock at redhat.com (Hugh Brock) Date: Tue, 13 Mar 2007 10:45:17 -0400 Subject: [et-mgmt-tools] libvirtError: virDomainCreateLinux() failed POST operation failed:, (xend.err "Error creating domain: (22, 'Invalid argument')") In-Reply-To: <45F6B760.7030602@gmail.com> References: <45F6B760.7030602@gmail.com> Message-ID: <45F6B8FD.6010901@redhat.com> rnijskens at gmail.com wrote: > Hi, > > Whatever I do, I get always this error: > > libvirtError: virDomainCreateLinux() failed POST operation failed: > (xend.err "Error creating domain: (22, 'Invalid argument')") Looks to me like you're trying to install Centos 4.4 as a paravirt guest. If I'm not mistaken, Centos 4.4 corresponds to RHEL 4 update 4, which does not support paravirtualization. Try Fedora Core 6, RHEL 4.5, or RHEL 5. --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From dhawal at netmagicsolutions.com Tue Mar 13 16:07:42 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Tue, 13 Mar 2007 21:37:42 +0530 Subject: [et-mgmt-tools] cft pre-requisites ruby(abi) and rhel4 Message-ID: <45F6CC4E.5070305@netmagicsolutions.com> David and List, I am trying to install cft on a test centos-4.4 (i386) server, using David's yum repo.. # yum --enablerepo=dlutter-rhel4 install cft Setting up Install Process Setting up repositories Reading repository metadata in from local files Parsing package install arguments Resolving Dependencies --> Populating transaction set with selected packages. Please wait. ---> Package cft.noarch 0:0.1.0-2.el4 set to be updated --> Running transaction check --> Processing Dependency: ruby(abi) = 1.8 for package: cft --> Processing Dependency: ruby-fam for package: cft --> Restarting Dependency Resolution with new changes. --> Populating transaction set with selected packages. Please wait. ---> Package ruby-fam.i386 0:0.2.0-4.el4 set to be updated --> Running transaction check --> Processing Dependency: ruby(abi) = 1.8 for package: cft --> Finished Dependency Resolution Error: Missing Dependency: ruby(abi) = 1.8 is needed by package cft Since ruby(abi) is not available for RHEL4, can you exclude it from the 'Requires'? here is small patch to fix the spec file. [root at other ~]# diff -u delme/cft.spec cft.spec --- delme/cft.spec 2007-01-26 06:56:41.000000000 +0530 +++ cft.spec 2007-03-13 21:33:44.000000000 +0530 @@ -1,5 +1,8 @@ %{!?ruby_sitelibdir: %define ruby_sitelibdir %(ruby -rrbconfig -e 'puts Config::CONFIG["sitelibdir"]')} +%define has_ruby_abi 0%{?fedora:%fedora} >= 5 || 0%{?rhel:%rhel} >= 5 +%define has_ruby_noarch %has_ruby_abi + Summary: Config file tracker Name: cft @@ -10,7 +13,10 @@ URL: http://cft.et.redhat.com/ Source: http://cft.et.redhat.com/download/cft-%{version}.tgz BuildRoot: %{_tmppath}/%{name}-%{version}-%{release}-root-%(%{__id_u} -n) +Requires: ruby >= 1.8.1 +%if %has_ruby_abi Requires: ruby ruby(abi) = 1.8 +%endif # FIXME: what version ? Requires: puppet Requires: ruby-fam - dhawal From hbrock at redhat.com Tue Mar 13 17:26:50 2007 From: hbrock at redhat.com (Hugh Brock) Date: Tue, 13 Mar 2007 13:26:50 -0400 Subject: [et-mgmt-tools] [PATCH] Check the making domain's mac address In-Reply-To: <45F66062.8010900@aa.jp.fujitsu.com> References: <45EE89B6.5010200@aa.jp.fujitsu.com> <45F66062.8010900@aa.jp.fujitsu.com> Message-ID: <45F6DEDA.7010001@redhat.com> Tatsuro Enokura wrote: > Hi, > > Would you give me a comment on this patch? > If not, please apply it. > > Tatsuro Enokura wrote: >> The virt-install command can specify the making domain's vnif >> MAC address. The MAC address must be unique on the system, >> but the virt-install command doesn't check that >> the MAC address is unique among the running domains and host. >> >> The attached patch resolve this issue in the following way: >> >> 1) Get the running Domain's vnif MAC address. >> 2) Get the host's NIC MAC address. >> 3) Check the making domain's MAC address with 1) and 2) 's data. > > Signed-off-by: Tatsuro Enokura > > Thanks, > Tatsuro Enokura > Hi there! Actually, you really don't need to build up a list of mac addresses and iterate over them. A much easier way is simply to get the XML for a domain and then use an xpath expression something like if ctx.xpathEval("count(/domain/devices/interface/mac/@address='%s')" % macaddr) > 0: # handle the case where the macaddr conflicts with an existing domain If you can rewrite the patch along these lines I'll be happy to take it. Thanks, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From dlutter at redhat.com Tue Mar 13 18:56:46 2007 From: dlutter at redhat.com (David Lutterkort) Date: Tue, 13 Mar 2007 11:56:46 -0700 Subject: [et-mgmt-tools] cft pre-requisites ruby(abi) and rhel4 In-Reply-To: <45F6CC4E.5070305@netmagicsolutions.com> References: <45F6CC4E.5070305@netmagicsolutions.com> Message-ID: <1173812206.17889.0.camel@localhost.localdomain> Hi Dhawal, On Tue, 2007-03-13 at 21:37 +0530, Dhawal Doshy wrote: > David and List, > > I am trying to install cft on a test centos-4.4 (i386) server, using > David's yum repo.. I just pushed an updated package to my yum repo (cft-0.1.0-3) ... that fixes the issue for me; if not let me know ;) David From ken at robson.net Tue Mar 13 20:14:33 2007 From: ken at robson.net (Ken Robson) Date: Tue, 13 Mar 2007 20:14:33 -0000 Subject: [et-mgmt-tools] Issues With $tree... In-Reply-To: <45F5634B.9020909@redhat.com> References: <010601c763fb$8d5186e0$0301a8c0@wxp1mbl1> <45F5634B.9020909@redhat.com> Message-ID: <003701c765ac$37ab1e10$0301a8c0@wxp1mbl1> > Thanks for the feedback... > > It looks like you were using "cobbler import --mirror" on a DVD > location, or perhaps the older "cobbler import --path". > Which was it? > > If it was "--mirror", that code, being designed for public > rsync mirrors > and not DVD's, likely did not have enough information to pick a good > kickstart for the distro you used. If it was "--path", that code > doesn't auto-assign kickstarts at all. Basically cobbler uses known > pieces of the path, which are found on all public mirrors, > like "FC-6" > or "RHEL-4" to determine what the distro is and to assign a good > kickstart to that distro. > > Either way, you can edit the profile file > (/var/lib/cobbler/profiles) to > add in a tree location to use. It's another element of > "ksmeta" and > just needs to have a value like > "http://server/cobbler_track/path/to/files/os" or equivalent. > > I'll go back and see if the kickstart assignment code can be > tweaked to > better suit the DVD import cases. I followed the section in the man page which was entitled "NORMAL WORKFLOW" which seemed to indicate that you should import a local DVD using: cobbler import --mirror=root at localhost:/mnt/dvd --mirror-name=any-name Or that was my interpretation! I will look at the man page more thoroughly and import the DVD using the --path option as you mention above. Thanks for your help. -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 3181 bytes Desc: not available URL: From ken at robson.net Tue Mar 13 20:17:50 2007 From: ken at robson.net (Ken Robson) Date: Tue, 13 Mar 2007 20:17:50 -0000 Subject: [et-mgmt-tools] Issues With $tree... In-Reply-To: <003701c765ac$37ab1e10$0301a8c0@wxp1mbl1> References: <010601c763fb$8d5186e0$0301a8c0@wxp1mbl1><45F5634B.9020909@redhat.com> <003701c765ac$37ab1e10$0301a8c0@wxp1mbl1> Message-ID: <004901c765ac$ae277430$0301a8c0@wxp1mbl1> > I followed the section in the man page which was entitled > "NORMAL WORKFLOW" > which seemed to indicate that you should import a local DVD using: Duh - sorry meant "IMPORT WORKFLOW" - after more thorough reading its clear to me I was using the wrong set of instructions. -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 3181 bytes Desc: not available URL: From mdehaan at redhat.com Tue Mar 13 19:31:23 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 13 Mar 2007 15:31:23 -0400 Subject: [et-mgmt-tools] Issues With $tree... In-Reply-To: <004901c765ac$ae277430$0301a8c0@wxp1mbl1> References: <010601c763fb$8d5186e0$0301a8c0@wxp1mbl1><45F5634B.9020909@redhat.com> <003701c765ac$37ab1e10$0301a8c0@wxp1mbl1> <004901c765ac$ae277430$0301a8c0@wxp1mbl1> Message-ID: <45F6FC0B.7090807@redhat.com> Ken Robson wrote: >> I followed the section in the man page which was entitled >> "NORMAL WORKFLOW" >> which seemed to indicate that you should import a local DVD using: >> > > Duh - sorry meant "IMPORT WORKFLOW" - after more thorough reading its clear > to me I was using the wrong set of instructions. > > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools That DVD import trick may have been added to the manpage too early... I'll look into the issues with tree and DVD imports soon. Until then, editing /var/lib/cobbler/profiles to add a "tree" parameter to "ksmeta" is a good route to go on. From berrange at redhat.com Tue Mar 13 21:27:02 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Tue, 13 Mar 2007 21:27:02 +0000 Subject: [et-mgmt-tools] About a solution to operate the virt-manager In-Reply-To: <20070307015845.GA9874@redhat.com> References: <45ED1D7E.9070805@jp.fujitsu.com> <20070306121553.GB21807@redhat.com> <45EE11D0.6070207@jp.fujitsu.com> <20070307015845.GA9874@redhat.com> Message-ID: <20070313212702.GA10124@redhat.com> On Wed, Mar 07, 2007 at 01:58:45AM +0000, Daniel P. Berrange wrote: > On Wed, Mar 07, 2007 at 10:13:52AM +0900, Yoshihiko Yaegashi wrote: > > Daniel P. Berrange wrote: > > > This is a GNOME / GTK bug I believe - try disabling 'Accessibility' > > > support in GNOME, and the logging out from X and strting a new > > > GNOME desktop session. > > > > Thank you for your advice. > > I have been possible to operate the virt-manager by the above method. > > > > I have a question. > > Is it possible to execute the Dogtail tests as 'Accessibility' > > support made disable? > > No, it is not possible - Dogtail uses the accessibility extensions to > control the application. Probably best to file a bug about the hang so > we can make sure it gets tracked & resolved in Fedora 7. In the meantime > things ought to be working correctly on FC6 with accessibility enabled. I've applied some patches to virt-manager which should resolve the problem. It is now possible to use it even with accessibility turned on. You need to have version 0.3.1-4.fc7 Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From ccoffing at novell.com Tue Mar 13 22:54:52 2007 From: ccoffing at novell.com (Charles Coffing) Date: Tue, 13 Mar 2007 16:54:52 -0600 Subject: [et-mgmt-tools] [PATCH] virt-manager: raising strings is deprecated Message-ID: <45F6D6A3.D169.003C.0@novell.com> Python has deprecated raising strings. The attached patch changes two such instances to instead raise a RuntimeError, to avoid warnings from python. Please apply. Signed-off-by: Charles Coffing -------------- next part -------------- A non-text attachment was scrubbed... Name: virtman-deprecated-raise.diff Type: application/octet-stream Size: 728 bytes Desc: not available URL: From ccoffing at novell.com Tue Mar 13 22:54:47 2007 From: ccoffing at novell.com (Charles Coffing) Date: Tue, 13 Mar 2007 16:54:47 -0600 Subject: [et-mgmt-tools] [PATCH] virt-manager: pass F10 through to guest Message-ID: <45F6D69E.D169.003C.0@novell.com> By default, F10 is bound to virt-manager's menu bar, so the guest can never get F10. The attached patch disables this accelerator for the virt-manager app when the keyboard is grabbed, and restores it when the keyboard is released. Please apply (or comment...) Signed-off-by: Charles Coffing -------------- next part -------------- A non-text attachment was scrubbed... Name: virtman-f10.diff Type: application/octet-stream Size: 924 bytes Desc: not available URL: From fred at x-tend.be Tue Mar 13 22:27:15 2007 From: fred at x-tend.be (=?ISO-8859-1?Q?Fr=E9d=E9ric?= Descamps) Date: Tue, 13 Mar 2007 23:27:15 +0100 Subject: [et-mgmt-tools] some cobbler questions Message-ID: <1173824835.6406.5.camel@localhost> Hello, I'd like to know if there is a way to make an installation of a domU but instead of using an xvda, use and lvm logical volume from the dom0 ? I really like to be able to create on my dom0 all the lvm that will be partitions for any of my domU. So for example I'll have on /dev/mapper VolGroup00-LogVolHome VolGroup00-LogVolVar VolGroup00-LogVolLib VolGroup00-LogVolUsr VolGroup00-LogVolSwap (all these for my dom0) VolGroup00-centos.root VolGroup00-centos.var (those for my centos domU) any idea how I could do that with cobbler ? (and if it's possible of course) so use something like 'phy:VolGroup00/centos.root,sda1,w' for my xen domU config thx, fred. -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From mdehaan at redhat.com Tue Mar 13 22:47:07 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 13 Mar 2007 18:47:07 -0400 Subject: [et-mgmt-tools] some cobbler questions In-Reply-To: <1173824835.6406.5.camel@localhost> References: <1173824835.6406.5.camel@localhost> Message-ID: <45F729EB.10900@redhat.com> Fr?d?ric Descamps wrote: > Hello, > > I'd like to know if there is a way to make an installation of a domU but > instead of using an xvda, use and lvm logical volume from the dom0 ? > > I really like to be able to create on my dom0 all the lvm that will be > partitions for any of my domU. > > So for example I'll have on /dev/mapper > VolGroup00-LogVolHome > VolGroup00-LogVolVar > VolGroup00-LogVolLib > VolGroup00-LogVolUsr > VolGroup00-LogVolSwap (all these for my dom0) > VolGroup00-centos.root > VolGroup00-centos.var (those for my centos domU) > > any idea how I could do that with cobbler ? (and if it's possible of > course) > > so use something like 'phy:VolGroup00/centos.root,sda1,w' for my xen > domU config > > > thx, > > fred. > > > > > > > Cobbler doesn't yet support installing into LVM partitions. Since koan 0.2.7 has just been moved over to use the "virtinst" library, though, it probably can be made to do this as long as virtinst supports that. (I believe it does). I'll put this on the investigate/TODO list, unless someone else wants to take a crack at it... --Michael From fred at x-tend.be Tue Mar 13 23:00:45 2007 From: fred at x-tend.be (=?ISO-8859-1?Q?Fr=E9d=E9ric?= Descamps) Date: Wed, 14 Mar 2007 00:00:45 +0100 Subject: [et-mgmt-tools] some cobbler questions Message-ID: <1173826845.8726.0.camel@localhost> Hello, I'd like to know if there is a way to make an installation of a domU but instead of using an xvda, use and lvm logical volume from the dom0 ? I really like to be able to create on my dom0 all the lvm that will be partitions for any of my domU. So for example I'll have on /dev/mapper VolGroup00-LogVolHome VolGroup00-LogVolVar VolGroup00-LogVolLib VolGroup00-LogVolUsr VolGroup00-LogVolSwap (all these for my dom0) VolGroup00-centos.root VolGroup00-centos.var (those for my centos domU) any idea how I could do that with cobbler ? (and if it's possible of course) so use something like 'phy:VolGroup00/centos.root,sda1,w' for my xen domU config thx, fred. -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From y.yaegashi at jp.fujitsu.com Wed Mar 14 05:07:17 2007 From: y.yaegashi at jp.fujitsu.com (Yoshihiko Yaegashi) Date: Wed, 14 Mar 2007 14:07:17 +0900 Subject: [et-mgmt-tools] About a solution to operate the virt-manager In-Reply-To: <20070313212702.GA10124@redhat.com> References: <45ED1D7E.9070805@jp.fujitsu.com> <20070306121553.GB21807@redhat.com> <45EE11D0.6070207@jp.fujitsu.com> <20070307015845.GA9874@redhat.com> <20070313212702.GA10124@redhat.com> Message-ID: <45F78305.70500@jp.fujitsu.com> Daniel P. Berrange wrote: > On Wed, Mar 07, 2007 at 01:58:45AM +0000, Daniel P. Berrange wrote: >> On Wed, Mar 07, 2007 at 10:13:52AM +0900, Yoshihiko Yaegashi wrote: >>> Daniel P. Berrange wrote: >>>> This is a GNOME / GTK bug I believe - try disabling 'Accessibility' >>>> support in GNOME, and the logging out from X and strting a new >>>> GNOME desktop session. >>> Thank you for your advice. >>> I have been possible to operate the virt-manager by the above method. >>> >>> I have a question. >>> Is it possible to execute the Dogtail tests as 'Accessibility' >>> support made disable? >> No, it is not possible - Dogtail uses the accessibility extensions to >> control the application. Probably best to file a bug about the hang so >> we can make sure it gets tracked & resolved in Fedora 7. In the meantime >> things ought to be working correctly on FC6 with accessibility enabled. > > I've applied some patches to virt-manager which should resolve the problem. > It is now possible to use it even with accessibility turned on. You need > to have version 0.3.1-4.fc7 Thank you for your reply. I want to try it. Can I have version 0.3.1-4.fc7? It seems that version 0.3.1-4.fc7 isn't uploaded to the server Thanks Yaegashi From dhawal at netmagicsolutions.com Wed Mar 14 06:32:32 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Wed, 14 Mar 2007 12:02:32 +0530 Subject: [et-mgmt-tools] cft pre-requisites ruby(abi) and rhel4 In-Reply-To: <1173812206.17889.0.camel@localhost.localdomain> References: <45F6CC4E.5070305@netmagicsolutions.com> <1173812206.17889.0.camel@localhost.localdomain> Message-ID: <45F79700.5020501@netmagicsolutions.com> David Lutterkort wrote: > Hi Dhawal, > > On Tue, 2007-03-13 at 21:37 +0530, Dhawal Doshy wrote: >> David and List, >> >> I am trying to install cft on a test centos-4.4 (i386) server, using >> David's yum repo.. > > I just pushed an updated package to my yum repo (cft-0.1.0-3) ... that > fixes the issue for me; if not let me know ;) i tried a yum install.... it worked as advertised.. will trouble you again once i actually start using it. thanks, - dhawal From fj7716hz at aa.jp.fujitsu.com Wed Mar 14 10:12:19 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Wed, 14 Mar 2007 19:12:19 +0900 Subject: [et-mgmt-tools] [PATCH] Check the making domain's mac address In-Reply-To: <45F6DEDA.7010001@redhat.com> References: <45EE89B6.5010200@aa.jp.fujitsu.com> <45F66062.8010900@aa.jp.fujitsu.com> <45F6DEDA.7010001@redhat.com> Message-ID: <45F7CA83.3020102@aa.jp.fujitsu.com> Hi, Hugh Hugh Brock wrote: >>> The virt-install command can specify the making domain's vnif >>> MAC address. The MAC address must be unique on the system, >>> but the virt-install command doesn't check that >>> the MAC address is unique among the running domains and host. >>> >>> The attached patch resolve this issue in the following way: >>> >>> 1) Get the running Domain's vnif MAC address. >>> 2) Get the host's NIC MAC address. >>> 3) Check the making domain's MAC address with 1) and 2) 's data. > Actually, you really don't need to build up a list of mac addresses and > iterate over them. A much easier way is simply to get the XML for a > domain and then use an xpath expression something like > > if ctx.xpathEval("count(/domain/devices/interface/mac/@address='%s')" % > macaddr) > 0: > # handle the case where the macaddr conflicts with an existing domain > > If you can rewrite the patch along these lines I'll be happy to take it. Thank you for your suggestion. I rewrite the patch. Thanks, Tatsuro Enokura ------------------------------------------------------------------------------- diff -r 6b63ce413aab virtinst/Guest.py --- a/virtinst/Guest.py Tue Mar 13 12:43:49 2007 -0400 +++ b/virtinst/Guest.py Wed Mar 14 18:08:31 2007 +0900 @@ -15,6 +15,7 @@ import os, os.path import os, os.path import stat, sys, time import re +import libxml2 import libvirt @@ -148,9 +149,76 @@ class VirtualNetworkInterface: self.macaddr = macaddr self.bridge = bridge - def setup(self): + def setup(self, conn): + # get Running Domains + ids = conn.listDomainsID(); + vms = [] + for id in ids: + vm = conn.lookupByID(id) + vms.append(vm) + + # get the Host's NIC MACaddress + hostdevs = util.get_host_network_devices() + + # check conflict MAC address if self.macaddr is None: - self.macaddr = util.randomMAC() + while 1: + self.macaddr = util.randomMAC() + for vm in vms: + doc = None + try: + doc = libxml2.parseDoc(vm.XMLDesc(0)) + except: + continue + ctx = doc.xpathNewContext() + try: + try: + count = ctx.xpathEval("count(/domain/devices/interface/mac[@address='%s'])" + % self.macaddr.upper()) + count += ctx.xpathEval("count(/domain/devices/interface/mac[@address='%s'])" + % self.macaddr.lower()) + if count > 0: + break + except: + continue + finally: + if ctx is not None: + ctx.xpathFreeContext() + if doc is not None: + doc.freeDoc() + else: + break + else: + for vm in vms: + doc = None + try: + doc = libxml2.parseDoc(vm.XMLDesc(0)) + except: + continue + ctx = doc.xpathNewContext() + try: + try: + count = ctx.xpathEval("count(/domain/devices/interface/mac[@address='%s'])" + % self.macaddr.upper()) + count += ctx.xpathEval("count(/domain/devices/interface/mac[@address='%s'])" + % self.macaddr.lower()) + if count > 0: + # conflict macaddr + raise RuntimeError, "The MAC address you entered is already in use by another guest!" + except RuntimeError: + raise + except: + continue + finally: + if ctx is not None: + ctx.xpathFreeContext() + if doc is not None: + doc.freeDoc() + for (dummy, dummy, dummy, dummy, host_macaddr) in hostdevs: + if self.macaddr.upper() == host_macaddr.upper(): + raise ValueError, "The MAC address you entered is conflict with the physical NIC." + + if not self.bridge: self.bridge = util.default_bridge() @@ -429,7 +497,7 @@ class Guest(object): for disk in self.disks: disk.setup(progresscb) for nic in self.nics: - nic.setup() + nic.setup(self.conn) def _get_disk_xml(self, install = True): """Get the disk config in the libvirt XML format""" diff -r 6b63ce413aab virtinst/util.py --- a/virtinst/util.py Tue Mar 13 12:43:49 2007 -0400 +++ b/virtinst/util.py Wed Mar 14 16:52:27 2007 +0900 @@ -123,3 +123,22 @@ def uuidFromString(s): def uuidFromString(s): s = s.replace('-', '') return [ int(s[i : i + 2], 16) for i in range(0, 32, 2) ] + +# the following function quotes from python2.5/uuid.py +def get_host_network_devices(): + device = [] + for dir in ['', '/sbin/', '/usr/sbin']: + executable = os.path.join(dir, "ifconfig") + if not os.path.exists(executable): + continue + try: + cmd = 'LC_ALL=C %s -a 2>/dev/null' % (executable) + pipe = os.popen(cmd) + except IOError: + continue + for line in pipe: + words = line.lower().split() + for i in range(len(words)): + if words[i] == "hwaddr": + device.append(words) + return device ------------------------------------------------------------------------------- From fj0873gn at aa.jp.fujitsu.com Wed Mar 14 10:19:46 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Wed, 14 Mar 2007 19:19:46 +0900 Subject: [et-mgmt-tools] [PATCH][RESEND] connecting console after installcompletes In-Reply-To: <45F6A3A8.7040708@redhat.com> References: <200703071836.BHE35429.K9O840GH@aa.jp.fujitsu.com> <200703131616.JGI57822.K09HG8O4@aa.jp.fujitsu.com> <45F6A3A8.7040708@redhat.com> Message-ID: <200703141919.GCJ39012.KH804GO9@aa.jp.fujitsu.com> Hi, Hugh > Two issues here: > > 1. The whole "restart the guest automatically" feature in virt-install > is a bit of a hack, and we are contemplating removing it altogether. The > problem is that virt-install has no real way of knowing whether the > guest that just shut down is really waiting to be rebooted, or is in > fact a failed install that the user would not want restarted anyway. Exactly. But, the "restart" hack is also true to help the user. Especially, the domain cannot be completely installed about Windows etc. with the reboot during the installation if it doesn't start automatically. Please keep autorestart function like --autorestart. > Having said this, as long as the "restart" hack is still there, I guess > I have no problem reconnecting the console as well. > > 2. Minor nit: instead of swallowing the error like > > + except OSError, (errno, msg): > + print __name__, "waitpid:", msg > > why not go ahead and raise a RuntimeError (and log the message) or > something similar? Okey. How about this correction? Thanks, Nobuhiro Itou. diff -r 6b63ce413aab virt-install --- a/virt-install Tue Mar 13 12:43:49 2007 -0400 +++ b/virt-install Wed Mar 14 06:16:00 2007 +0900 @@ -517,6 +517,7 @@ def main(): # the domain print "Guest installation complete... restarting guest." dom.create() + guest.connect_console(conscb) else: print ("Domain installation does not appear to have been\n" "successful. If it was, you can restart your domain\n" diff -r 6b63ce413aab virtinst/Guest.py --- a/virtinst/Guest.py Tue Mar 13 12:43:49 2007 -0400 +++ b/virtinst/Guest.py Wed Mar 14 15:35:56 2007 +0900 @@ -575,6 +575,18 @@ class Guest(object): # for inactive guest, or get the still running install.. return self.conn.lookupByName(self.name) + def connect_console(self, consolecb): + child = None + if consolecb: + logging.debug("Launching console callback") + child = consolecb(self.domain) + + if child: # if we connected the console, wait for it to finish + try: + (pid, status) = os.waitpid(child, 0) + except OSError, (errno, msg): + raise RuntimeError, "waiting console pid error: %s" % msg + def validate_parms(self): if self.domain is not None: raise RuntimeError, "Domain already started!" From berrange at redhat.com Wed Mar 14 11:33:04 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 14 Mar 2007 11:33:04 +0000 Subject: [et-mgmt-tools] About a solution to operate the virt-manager In-Reply-To: <45F78305.70500@jp.fujitsu.com> References: <45ED1D7E.9070805@jp.fujitsu.com> <20070306121553.GB21807@redhat.com> <45EE11D0.6070207@jp.fujitsu.com> <20070307015845.GA9874@redhat.com> <20070313212702.GA10124@redhat.com> <45F78305.70500@jp.fujitsu.com> Message-ID: <20070314113304.GB29712@redhat.com> On Wed, Mar 14, 2007 at 02:07:17PM +0900, Yoshihiko Yaegashi wrote: > Daniel P. Berrange wrote: > >On Wed, Mar 07, 2007 at 01:58:45AM +0000, Daniel P. Berrange wrote: > >>On Wed, Mar 07, 2007 at 10:13:52AM +0900, Yoshihiko Yaegashi wrote: > >>>Daniel P. Berrange wrote: > >>>>This is a GNOME / GTK bug I believe - try disabling 'Accessibility' > >>>>support in GNOME, and the logging out from X and strting a new > >>>>GNOME desktop session. > >>>Thank you for your advice. > >>>I have been possible to operate the virt-manager by the above method. > >>> > >>>I have a question. > >>>Is it possible to execute the Dogtail tests as 'Accessibility' > >>>support made disable? > >>No, it is not possible - Dogtail uses the accessibility extensions to > >>control the application. Probably best to file a bug about the hang so > >>we can make sure it gets tracked & resolved in Fedora 7. In the meantime > >>things ought to be working correctly on FC6 with accessibility enabled. > > > >I've applied some patches to virt-manager which should resolve the problem. > >It is now possible to use it even with accessibility turned on. You need > >to have version 0.3.1-4.fc7 > > Thank you for your reply. > I want to try it. > Can I have version 0.3.1-4.fc7? > It seems that version 0.3.1-4.fc7 isn't uploaded to the server It is on the server now - there must have beeen a delay with the push to rawhide from the build system http://download.fedora.redhat.com/pub/fedora/linux/core/development/x86_64/os/Fedora/virt-manager-0.3.1-4.fc7.x86_64.rpm Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From sakaia at jp.fujitsu.com Wed Mar 14 12:21:09 2007 From: sakaia at jp.fujitsu.com (Atsushi SAKAI) Date: Wed, 14 Mar 2007 21:21:09 +0900 Subject: [et-mgmt-tools] About a solution to operate the virt-manager In-Reply-To: <20070314113304.GB29712@redhat.com> References: <20070314113304.GB29712@redhat.com> Message-ID: <200703141221.l2ECLXka001112@fjmscan502.ms.jp.fujitsu.com> Hi, Dan I confirmed x86_64 update as you suggested but not found for i386. like this http://download.fedora.redhat.com/pub/fedora/linux/core/development/i386/os/Fedora/virt-manager-0.3.1-4.fc7.i386.rpm Is there 0.3.1-4 for i386 is available? This is just a time delay effect of the fedora server? Thanks Atsushi SAKAI > > >I've applied some patches to virt-manager which should resolve the problem. > > >It is now possible to use it even with accessibility turned on. You need > > >to have version 0.3.1-4.fc7 > > > > Thank you for your reply. > > I want to try it. > > Can I have version 0.3.1-4.fc7? > > It seems that version 0.3.1-4.fc7 isn't uploaded to the server > > It is on the server now - there must have beeen a delay with the push to > rawhide from the build system > > http://download.fedora.redhat.com/pub/fedora/linux/core/development/x86_64/os/Fedora/virt-manager-0.3.1-4.fc7.x86_64.rpm > > Dan. > -- > |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| > |=- Perl modules: http://search.cpan.org/~danberr/ -=| > |=- Projects: http://freshmeat.net/~danielpb/ -=| > |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From berrange at redhat.com Wed Mar 14 12:37:50 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 14 Mar 2007 12:37:50 +0000 Subject: [et-mgmt-tools] About a solution to operate the virt-manager In-Reply-To: <200703141221.l2ECLXka001112@fjmscan502.ms.jp.fujitsu.com> References: <20070314113304.GB29712@redhat.com> <200703141221.l2ECLXka001112@fjmscan502.ms.jp.fujitsu.com> Message-ID: <20070314123750.GD29712@redhat.com> On Wed, Mar 14, 2007 at 09:21:09PM +0900, Atsushi SAKAI wrote: > Hi, Dan > > I confirmed x86_64 update as you suggested but not found for i386. > like this > http://download.fedora.redhat.com/pub/fedora/linux/core/development/i386/os/Fedora/virt-manager-0.3.1-4.fc7.i386.rpm > Is there 0.3.1-4 for i386 is available? > This is just a time delay effect of the fedora server? The i386 link you show above is working fine for me. Perhaps it was just a time delay syncing. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From hbrock at redhat.com Wed Mar 14 20:08:35 2007 From: hbrock at redhat.com (Hugh Brock) Date: Wed, 14 Mar 2007 16:08:35 -0400 Subject: [et-mgmt-tools] [PATCH] virt-manager: pass F10 through to guest In-Reply-To: <45F6D69E.D169.003C.0@novell.com> References: <45F6D69E.D169.003C.0@novell.com> Message-ID: <45F85643.7090508@redhat.com> Charles Coffing wrote: > By default, F10 is bound to virt-manager's menu bar, so the guest can never get F10. > > The attached patch disables this accelerator for the virt-manager app when the keyboard is grabbed, and restores it when the keyboard is released. > > Please apply (or comment...) > > Signed-off-by: Charles Coffing > > > Applied, thanks! --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From hbrock at redhat.com Wed Mar 14 20:08:48 2007 From: hbrock at redhat.com (Hugh Brock) Date: Wed, 14 Mar 2007 16:08:48 -0400 Subject: [et-mgmt-tools] [PATCH] virt-manager: raising strings is deprecated In-Reply-To: <45F6D6A3.D169.003C.0@novell.com> References: <45F6D6A3.D169.003C.0@novell.com> Message-ID: <45F85650.5010903@redhat.com> Charles Coffing wrote: > Python has deprecated raising strings. The attached patch changes two such instances to instead raise a RuntimeError, to avoid warnings from python. Please apply. > > Signed-off-by: Charles Coffing > Applied, thanks! --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From mdehaan at redhat.com Wed Mar 14 20:45:39 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 14 Mar 2007 16:45:39 -0400 Subject: [et-mgmt-tools] Koan 0.2.7 released (WAS: Greatly improved Xen support for koan) In-Reply-To: <45F09B1D.2010007@redhat.com> References: <45F09B1D.2010007@redhat.com> Message-ID: <45F85EF3.6070609@redhat.com> Michael DeHaan wrote: > For those using Cobbler for virtualization installs, I've commited a > change to the Mercurial repository that you will probably be > interested in. > Basically koan now uses the "virtinst" library to install virtualized > images rather than using "libvirt" directly. "virtinst" is the same > library that virt-manager uses to install virtual machines. So, how > does this affect the running of koan? > > -- The biggest change is that koan now terminates very quickly, with > the rest of the virtual installation going on in the background. > This allows multiple parallel virtual installs if someone wants to do > it. This should make koan much more usable in batch context. If > you want to watch the installation progress, usage of xm or > virt-manager will allow that). > -- Koan also names each virtual system after the mac address they are > created with, such that they are unique. So no more "xenguest" and > "xenguest_1" kind of thing, which was error prone when you installed a > machine and then it crashed, causing naming conflicts. (--xen-name > will probably be removed from cobbler also, as it's no longer > needed). If you are using koan with "--system" instead of > "--profile", the MAC address can be controlled from cobbler, along > with other fine tunning arguments. If not, it will be randomly > generated. If anyone has any objections to this naming scheme, you > might want to speak up. I can think of a couple of alternate ways to > name domains automatically -- "$profilename_$timestamp" isn't so bad. > -- Koan no longer writes any files to /var/koan -- anywhere, because > it doesn't have to. > This is still undergoing some testing, but I would call it very stable > at this point. I'm just going to be tweaking output and error > handling a bit, if anything. > Further down the road, this will enable koan to do hardware virtual > installs and handle other virtualization types. > Note that depending on what distro you are running, Xen will save the > configuration files in one of the following: > /etc/xen > /var/lib/xen > /var/run/xen > > So, if you have domains that get stopped, looking in those directories > will tell you which ones you have to start. Also on FC6 and later > you can experiment with "virsh" for management of inactive domains. > > 0.2.7 will probably be released sometime later next week ... unless > you want to play with the upstream code (please do, not too many > computers have exploded from using it, I promise) > > --Michael > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools Koan 0.2.7 (with the above virtualization changes) is now released. I've used it quite a bit and it seems all stable and happy. If you find problems, let me know. Pushed to the Fedora mirrors, otherwise grab the RPMs from cobbler.et.redhat.com or build from source. Thanks, --Michael From hbrock at redhat.com Wed Mar 14 21:11:19 2007 From: hbrock at redhat.com (Hugh Brock) Date: Wed, 14 Mar 2007 17:11:19 -0400 Subject: [et-mgmt-tools] [PATCH][RESEND] connecting console after installcompletes In-Reply-To: <200703141919.GCJ39012.KH804GO9@aa.jp.fujitsu.com> References: <200703071836.BHE35429.K9O840GH@aa.jp.fujitsu.com> <200703131616.JGI57822.K09HG8O4@aa.jp.fujitsu.com> <45F6A3A8.7040708@redhat.com> <200703141919.GCJ39012.KH804GO9@aa.jp.fujitsu.com> Message-ID: <45F864F7.7070200@redhat.com> Nobuhiro Itou wrote: > Hi, Hugh > >> Two issues here: >> >> 1. The whole "restart the guest automatically" feature in virt-install >> is a bit of a hack, and we are contemplating removing it altogether. The >> problem is that virt-install has no real way of knowing whether the >> guest that just shut down is really waiting to be rebooted, or is in >> fact a failed install that the user would not want restarted anyway. > > Exactly. > But, the "restart" hack is also true to help the user. > Especially, the domain cannot be completely installed about Windows etc. > with the reboot during the installation if it doesn't start automatically. > Please keep autorestart function like --autorestart. > >> Having said this, as long as the "restart" hack is still there, I guess >> I have no problem reconnecting the console as well. >> >> 2. Minor nit: instead of swallowing the error like >> >> + except OSError, (errno, msg): >> + print __name__, "waitpid:", msg >> >> why not go ahead and raise a RuntimeError (and log the message) or >> something similar? > > Okey. > How about this correction? > > > Thanks, > Nobuhiro Itou. > Thanks! I am working on fixing the auto-restart feature now, the console connection part isn't working properly at the moment. Once I have fixed it, I'll apply your patch as well. Take care, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From hbrock at redhat.com Wed Mar 14 21:14:30 2007 From: hbrock at redhat.com (Hugh Brock) Date: Wed, 14 Mar 2007 17:14:30 -0400 Subject: [et-mgmt-tools] [PATCH] Check the making domain's mac address In-Reply-To: <45F7CA83.3020102@aa.jp.fujitsu.com> References: <45EE89B6.5010200@aa.jp.fujitsu.com> <45F66062.8010900@aa.jp.fujitsu.com> <45F6DEDA.7010001@redhat.com> <45F7CA83.3020102@aa.jp.fujitsu.com> Message-ID: <45F865B6.9090902@redhat.com> Tatsuro Enokura wrote: > Hi, Hugh > > Hugh Brock wrote: >>>> The virt-install command can specify the making domain's vnif >>>> MAC address. The MAC address must be unique on the system, >>>> but the virt-install command doesn't check that >>>> the MAC address is unique among the running domains and host. >>>> >>>> The attached patch resolve this issue in the following way: >>>> >>>> 1) Get the running Domain's vnif MAC address. >>>> 2) Get the host's NIC MAC address. >>>> 3) Check the making domain's MAC address with 1) and 2) 's data. > >> Actually, you really don't need to build up a list of mac addresses and >> iterate over them. A much easier way is simply to get the XML for a >> domain and then use an xpath expression something like >> >> if >> ctx.xpathEval("count(/domain/devices/interface/mac/@address='%s')" % >> macaddr) > 0: >> # handle the case where the macaddr conflicts with an existing domain >> >> If you can rewrite the patch along these lines I'll be happy to take it. > > Thank you for your suggestion. > I rewrite the patch. > > Thanks, > Tatsuro Enokura I'm reviewing this now. I'm trying to work out if there's a way to do it without repeating so much code... :)... Thanks, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From hbrock at redhat.com Wed Mar 14 21:24:25 2007 From: hbrock at redhat.com (Hugh Brock) Date: Wed, 14 Mar 2007 17:24:25 -0400 Subject: [et-mgmt-tools] [PATCH][RESEND] connecting console after installcompletes In-Reply-To: <200703141919.GCJ39012.KH804GO9@aa.jp.fujitsu.com> References: <200703071836.BHE35429.K9O840GH@aa.jp.fujitsu.com> <200703131616.JGI57822.K09HG8O4@aa.jp.fujitsu.com> <45F6A3A8.7040708@redhat.com> <200703141919.GCJ39012.KH804GO9@aa.jp.fujitsu.com> Message-ID: <45F86809.2070109@redhat.com> Nobuhiro Itou wrote: > Hi, Hugh > >> Two issues here: >> >> 1. The whole "restart the guest automatically" feature in virt-install >> is a bit of a hack, and we are contemplating removing it altogether. The >> problem is that virt-install has no real way of knowing whether the >> guest that just shut down is really waiting to be rebooted, or is in >> fact a failed install that the user would not want restarted anyway. > > Exactly. > But, the "restart" hack is also true to help the user. > Especially, the domain cannot be completely installed about Windows etc. > with the reboot during the installation if it doesn't start automatically. > Please keep autorestart function like --autorestart. > OK, this is working fine now, it was a testing issue on my system. I have applied the patch. Thanks! --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From fred at x-tend.be Wed Mar 14 21:33:17 2007 From: fred at x-tend.be (=?ISO-8859-1?Q?Fr=E9d=E9ric?= Descamps) Date: Wed, 14 Mar 2007 22:33:17 +0100 Subject: [et-mgmt-tools] some cobbler questions In-Reply-To: <45F729EB.10900@redhat.com> References: <1173824835.6406.5.camel@localhost> <45F729EB.10900@redhat.com> Message-ID: <1173907997.4323.14.camel@localhost> On Tue, 2007-03-13 at 18:47 -0400, Michael DeHaan wrote: > Fr?d?ric Descamps wrote: > > Hello, > > > > I'd like to know if there is a way to make an installation of a domU but > > instead of using an xvda, use and lvm logical volume from the dom0 ? > > > > I really like to be able to create on my dom0 all the lvm that will be > > partitions for any of my domU. > > > > So for example I'll have on /dev/mapper > > VolGroup00-LogVolHome > > VolGroup00-LogVolVar > > VolGroup00-LogVolLib > > VolGroup00-LogVolUsr > > VolGroup00-LogVolSwap (all these for my dom0) > > VolGroup00-centos.root > > VolGroup00-centos.var (those for my centos domU) > > > > any idea how I could do that with cobbler ? (and if it's possible of > > course) > > > > so use something like 'phy:VolGroup00/centos.root,sda1,w' for my xen > > domU config > > > > > > thx, > > > > fred. > > > > > > > > > > > > > > > Cobbler doesn't yet support installing into LVM partitions. Since > koan 0.2.7 has just been moved over to use the "virtinst" library, > though, it probably can be made to do this as long as virtinst supports > that. (I believe it does). > > I'll put this on the investigate/TODO list, unless someone else wants to > take a crack at it... > > --Michael > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > Michael, I hope your todo list is not too long ;) My knowledge in python are limited... so it won't be easy for me to hack it :( regards, fred. -- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. From y.yaegashi at jp.fujitsu.com Thu Mar 15 04:42:01 2007 From: y.yaegashi at jp.fujitsu.com (Yoshihiko Yaegashi) Date: Thu, 15 Mar 2007 13:42:01 +0900 Subject: [et-mgmt-tools] About a solution to operate the virt-manager In-Reply-To: <20070313212702.GA10124@redhat.com> References: <45ED1D7E.9070805@jp.fujitsu.com> <20070306121553.GB21807@redhat.com> <45EE11D0.6070207@jp.fujitsu.com> <20070307015845.GA9874@redhat.com> <20070313212702.GA10124@redhat.com> Message-ID: <45F8CE99.7010100@jp.fujitsu.com> Daniel P. Berrange wrote: > I've applied some patches to virt-manager which should resolve the problem. > It is now possible to use it even with accessibility turned on. You need > to have version 0.3.1-4.fc7 Hi, Dan I have installed 0.3.1-4.fc7 for i386, but it is impossible to use virt-manager with accessibility turned on. Thanks Yaegashi From fj7716hz at aa.jp.fujitsu.com Thu Mar 15 08:45:26 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Thu, 15 Mar 2007 17:45:26 +0900 Subject: [et-mgmt-tools] [PATCH] Check the making domain's disk sources Message-ID: <45F907A6.6070100@aa.jp.fujitsu.com> Hi, The virt-install command specify the Domain's disk sources. But the virt-install command doesn't check that the disk sources are conflicted among the other domains. The attached patch resolve this issue in the following way: 1) Count that the making domian's disk src is already use the exists domain's disk source. 2) if count > 0, print confirmation message. 3) if type yes then leave disk source as it is, else ask disk source again. I am making a patch for virt-manager about this issue. Signed-off-by: Tatsuro Enokura Thanks, Tatsuro Enokura. -------------------------------------------------------------------- diff -r 057e8c1b54df virt-install --- a/virt-install Wed Mar 14 16:11:43 2007 -0400 +++ b/virt-install Thu Mar 15 16:31:43 2007 +0900 @@ -103,7 +103,7 @@ def get_keymap(keymap, guest): except ValueError, e: print "ERROR: ", e -def get_disk(disk, size, sparse, guest, hvm): +def get_disk(disk, size, sparse, guest, hvm, conn): # FIXME: need to handle a list of disks at some point while 1: msg = "What would you like to use as the disk (path)?" @@ -121,6 +121,57 @@ def get_disk(disk, size, sparse, guest, print "ERROR: ", e size = None + # disk src confilct check + vms = [] + # get working domain's name + ids = conn.listDomainsID(); + for id in ids: + vm = conn.lookupByID(id) + vms.append(vm) + # get defined domain + names = conn.listDefinedDomains() + for name in names: + vm = conn.lookupByName(name) + vms.append(vm) + + count = 0 + for vm in vms: + doc = None + try: + doc = libxml2.parseDoc(vm.XMLDesc(0)) + except: + continue + ctx = doc.xpathNewContext() + try: + try: + count += ctx.xpathEval("count(/domain/devices/disk/source[@dev='%s'])" % disk) + except: + continue + finally: + if ctx is not None: + ctx.xpathFreeContext() + if doc is not None: + doc.freeDoc() + if count > 0: + # confilct disk src + while 1: + retryFlg = False + warnmsg = "Disk %s is already in use by another guest!" % disk + res = prompt_for_input(warnmsg + " Do you really want to use the disk (yes or no)? ") + try: + if yes_or_no(res) is True: + break + else: + retryFlg = True + break + except ValueError, e: + print "ERROR: ", e + continue + if retryFlg is True: + # disk data clear, ask agein + disk = None + continue + try: d = virtinst.VirtualDisk(disk, size, sparse = sparse) if d.type == virtinst.VirtualDisk.TYPE_FILE and not(hvm) and virtinst.util.is_blktap_capable(): @@ -133,7 +184,7 @@ def get_disk(disk, size, sparse, guest, guest.disks.append(d) break -def get_disks(disk, size, sparse, guest, hvm): +def get_disks(disk, size, sparse, guest, hvm, conn): # ensure we have equal length lists if (type(disk) == type(size) == list): if len(disk) != len(size): @@ -145,13 +196,13 @@ def get_disks(disk, size, sparse, guest, disk = [ None ] * len(size) if (type(disk) == list): - map(lambda d, s: get_disk(d, s, sparse, guest, hvm), + map(lambda d, s: get_disk(d, s, sparse, guest, hvm, conn), disk, size) elif (type(size) == list): - map(lambda d, s: get_disk(d, s, sparse, guest, hvm), + map(lambda d, s: get_disk(d, s, sparse, guest, hvm, conn), disk, size) else: - get_disk(disk, size, sparse, guest, hvm) + get_disk(disk, size, sparse, guest, hvm, conn) def get_network(mac, bridge, guest): if mac == "RANDOM": @@ -461,7 +512,7 @@ def main(): # set up disks get_disks(options.diskfile, options.disksize, options.sparse, - guest, hvm) + guest, hvm, conn) # set up network information get_networks(options.mac, options.bridge, guest) -------------------------------------------------------------------- From fj7716hz at aa.jp.fujitsu.com Thu Mar 15 10:52:16 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Thu, 15 Mar 2007 19:52:16 +0900 Subject: [et-mgmt-tools] [PATCH] Check the making domain's mac address In-Reply-To: <45F865B6.9090902@redhat.com> References: <45EE89B6.5010200@aa.jp.fujitsu.com> <45F66062.8010900@aa.jp.fujitsu.com> <45F6DEDA.7010001@redhat.com> <45F7CA83.3020102@aa.jp.fujitsu.com> <45F865B6.9090902@redhat.com> Message-ID: <45F92560.80005@aa.jp.fujitsu.com> Hi, Hugh Hugh Brock wrote: >>>>> The virt-install command can specify the making domain's vnif >>>>> MAC address. The MAC address must be unique on the system, >>>>> but the virt-install command doesn't check that >>>>> the MAC address is unique among the running domains and host. >>>>> >>>>> The attached patch resolve this issue in the following way: >>>>> >>>>> 1) Get the running Domain's vnif MAC address. >>>>> 2) Get the host's NIC MAC address. >>>>> 3) Check the making domain's MAC address with 1) and 2) 's data. >> >>> Actually, you really don't need to build up a list of mac addresses and >>> iterate over them. A much easier way is simply to get the XML for a >>> domain and then use an xpath expression something like >>> >>> if >>> ctx.xpathEval("count(/domain/devices/interface/mac/@address='%s')" % >>> macaddr) > 0: >>> # handle the case where the macaddr conflicts with an existing >>> domain >>> >>> If you can rewrite the patch along these lines I'll be happy to take it. >> >> Thank you for your suggestion. >> I rewrite the patch. > > I'm reviewing this now. I'm trying to work out if there's a way to do it > without repeating so much code... :)... Thank you for viewing my code. I rewrite the patch that the repeating code become the method. Thanks, Tatsuro Enokura -------------------------------------------------------------------- diff -r 057e8c1b54df virtinst/Guest.py --- a/virtinst/Guest.py Wed Mar 14 16:11:43 2007 -0400 +++ b/virtinst/Guest.py Thu Mar 15 15:07:26 2007 +0900 @@ -15,6 +15,7 @@ import os, os.path import os, os.path import stat, sys, time import re +import libxml2 import libvirt @@ -148,9 +149,32 @@ class VirtualNetworkInterface: self.macaddr = macaddr self.bridge = bridge - def setup(self): + def setup(self, conn): + # get Running Domains + ids = conn.listDomainsID(); + vms = [] + for id in ids: + vm = conn.lookupByID(id) + vms.append(vm) + + # get the Host's NIC MACaddress + hostdevs = util.get_host_network_devices() + + # check conflict MAC address if self.macaddr is None: - self.macaddr = util.randomMAC() + while 1: + self.macaddr = util.randomMAC() + if self.countMACaddr(vms) > 0: + continue + else: + break + else: + if self.countMACaddr(vms) > 0: + raise RuntimeError, "The MAC address you entered is already in use by another guest!" + for (dummy, dummy, dummy, dummy, host_macaddr) in hostdevs: + if self.macaddr.upper() == host_macaddr.upper(): + raise RuntimeError, "The MAC address you entered is conflict with the physical NIC." + if not self.bridge: self.bridge = util.default_bridge() @@ -160,6 +184,30 @@ class VirtualNetworkInterface: " \n" + \ " \n") % \ { "bridge": self.bridge, "mac": self.macaddr } + + def countMACaddr(self, vms): + count = 0 + for vm in vms: + doc = None + try: + doc = libxml2.parseDoc(vm.XMLDesc(0)) + except: + continue + ctx = doc.xpathNewContext() + try: + try: + count += ctx.xpathEval("count(/domain/devices/interface/mac[@address='%s'])" + % self.macaddr.upper()) + count += ctx.xpathEval("count(/domain/devices/interface/mac[@address='%s'])" + % self.macaddr.lower()) + except: + continue + finally: + if ctx is not None: + ctx.xpathFreeContext() + if doc is not None: + doc.freeDoc() + return count # Back compat class to avoid ABI break class XenNetworkInterface(VirtualNetworkInterface): @@ -429,7 +477,7 @@ class Guest(object): for disk in self.disks: disk.setup(progresscb) for nic in self.nics: - nic.setup() + nic.setup(self.conn) def _get_disk_xml(self, install = True): """Get the disk config in the libvirt XML format""" diff -r 057e8c1b54df virtinst/util.py --- a/virtinst/util.py Wed Mar 14 16:11:43 2007 -0400 +++ b/virtinst/util.py Thu Mar 15 14:05:00 2007 +0900 @@ -123,3 +123,22 @@ def uuidFromString(s): def uuidFromString(s): s = s.replace('-', '') return [ int(s[i : i + 2], 16) for i in range(0, 32, 2) ] + +# the following function quotes from python2.5/uuid.py +def get_host_network_devices(): + device = [] + for dir in ['', '/sbin/', '/usr/sbin']: + executable = os.path.join(dir, "ifconfig") + if not os.path.exists(executable): + continue + try: + cmd = 'LC_ALL=C %s -a 2>/dev/null' % (executable) + pipe = os.popen(cmd) + except IOError: + continue + for line in pipe: + words = line.lower().split() + for i in range(len(words)): + if words[i] == "hwaddr": + device.append(words) + return device -------------------------------------------------------------------- From fj0873gn at aa.jp.fujitsu.com Thu Mar 15 11:13:19 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Thu, 15 Mar 2007 20:13:19 +0900 Subject: [et-mgmt-tools] [PATCH] Fix paravirt install by NFS Message-ID: <200703152013.DFG04184.0G98OHK4@aa.jp.fujitsu.com> Hi, The paravirt domain cannot be installed by NFS. The attached patch adds to solve this problem. Signed-off-by: Nobuhiro Itou Thanks, Nobuhiro Itou. diff -r 057e8c1b54df virtinst/DistroManager.py --- a/virtinst/DistroManager.py Wed Mar 14 16:11:43 2007 -0400 +++ b/virtinst/DistroManager.py Thu Mar 15 19:21:09 2007 +0900 @@ -101,10 +101,10 @@ class MountedImageFetcher(ImageFetcher): cmd = ["mount", "-o", "ro", self.location, self.mntdir] else: cmd = ["mount", "-o", "ro,loop", self.location, self.mntdir] - ret = subprocess.call(cmd) - if ret != 0: - self.cleanupLocation() - return False + ret = subprocess.call(cmd) + if ret != 0: + self.cleanupLocation() + return False return True def cleanupLocation(self): @@ -123,7 +123,7 @@ class MountedImageFetcher(ImageFetcher): base = os.path.basename(filename) try: src = self.mntdir + "/" + filename - if stat.S_ISDIR(os.stat(src)): + if stat.S_ISDIR(os.stat(src)[stat.ST_MODE]): pass else: file = open(src, "r") From berrange at redhat.com Thu Mar 15 12:15:04 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 15 Mar 2007 12:15:04 +0000 Subject: [et-mgmt-tools] [PATCH] Check the making domain's disk sources In-Reply-To: <45F907A6.6070100@aa.jp.fujitsu.com> References: <45F907A6.6070100@aa.jp.fujitsu.com> Message-ID: <20070315121503.GA1253@redhat.com> On Thu, Mar 15, 2007 at 05:45:26PM +0900, Tatsuro Enokura wrote: > Hi, > > The virt-install command specify the Domain's disk sources. > But the virt-install command doesn't check that the disk sources are > conflicted among the other domains. > > The attached patch resolve this issue in the following way: > > 1) Count that the making domian's disk src is already use > the exists domain's disk source. > 2) if count > 0, print confirmation message. > 3) if type yes then leave disk source as it is, > else ask disk source again. > > I am making a patch for virt-manager about this issue. If you the non-interactive bit of code into the virtinst/Guest.py module in the VirtualDisk class, we can share 95% of it between both virtinstall & virt-manager. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From hbrock at redhat.com Thu Mar 15 15:42:11 2007 From: hbrock at redhat.com (Hugh Brock) Date: Thu, 15 Mar 2007 11:42:11 -0400 Subject: [et-mgmt-tools] [PATCH] Check the making domain's mac address In-Reply-To: <45F92560.80005@aa.jp.fujitsu.com> References: <45EE89B6.5010200@aa.jp.fujitsu.com> <45F66062.8010900@aa.jp.fujitsu.com> <45F6DEDA.7010001@redhat.com> <45F7CA83.3020102@aa.jp.fujitsu.com> <45F865B6.9090902@redhat.com> <45F92560.80005@aa.jp.fujitsu.com> Message-ID: <45F96953.1010001@redhat.com> Tatsuro Enokura wrote: > Hi, Hugh > > Hugh Brock wrote: >>>>>> The virt-install command can specify the making domain's vnif >>>>>> MAC address. The MAC address must be unique on the system, >>>>>> but the virt-install command doesn't check that >>>>>> the MAC address is unique among the running domains and host. >>>>>> >>>>>> The attached patch resolve this issue in the following way: >>>>>> >>>>>> 1) Get the running Domain's vnif MAC address. >>>>>> 2) Get the host's NIC MAC address. >>>>>> 3) Check the making domain's MAC address with 1) and 2) 's data. >>>> Actually, you really don't need to build up a list of mac addresses and >>>> iterate over them. A much easier way is simply to get the XML for a >>>> domain and then use an xpath expression something like >>>> >>>> if >>>> ctx.xpathEval("count(/domain/devices/interface/mac/@address='%s')" % >>>> macaddr) > 0: >>>> # handle the case where the macaddr conflicts with an existing >>>> domain >>>> >>>> If you can rewrite the patch along these lines I'll be happy to take it. >>> Thank you for your suggestion. >>> I rewrite the patch. >> I'm reviewing this now. I'm trying to work out if there's a way to do it >> without repeating so much code... :)... > > Thank you for viewing my code. > I rewrite the patch that the repeating code become the method. > > Thanks, > Tatsuro Enokura That looks much better. I've tested and committed it. Thanks! Take care, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From hbrock at redhat.com Thu Mar 15 16:19:04 2007 From: hbrock at redhat.com (Hugh Brock) Date: Thu, 15 Mar 2007 12:19:04 -0400 Subject: [et-mgmt-tools] [PATCH] Fix paravirt install by NFS In-Reply-To: <200703152013.DFG04184.0G98OHK4@aa.jp.fujitsu.com> References: <200703152013.DFG04184.0G98OHK4@aa.jp.fujitsu.com> Message-ID: <45F971F8.7080305@redhat.com> Nobuhiro Itou wrote: > Hi, > > The paravirt domain cannot be installed by NFS. > The attached patch adds to solve this problem. > > Signed-off-by: Nobuhiro Itou > > > Thanks, > Nobuhiro Itou. > Applied, thanks very much! Take care, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From mdehaan at redhat.com Thu Mar 15 23:05:29 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 15 Mar 2007 19:05:29 -0400 Subject: [et-mgmt-tools] Announcing virt-factory 0.0.1 Message-ID: <45F9D139.1060307@redhat.com> Today we're pleased to announce the first public release of a new systems management project -- virt-factory -- currently at bright, new, and shiny version 0.0.1. Virt-factory is a project that aims to manage very large numbers of virtual machines in a very managable way. As it's a very new project, there is a lot of room for developers who are interested to get involved. We want this to be very much a community project. Check out the web site at http://virt-factory.et.redhat.com for more info, and perhaps also look at the Roadmap (http://et.redhat.com/page/VF_Roadmap) to see where we'd like to take it. === From the website: Virt-Factory manages virtualized infrastructure: * it focuses on interacting with large numbers of virtual systems and on addressing some of the interaction problems that brings with it * it is primarily aimed at a fairly formal setting (data center), though we hope it is useful on smaller scales, too * even though it has some uses for bare-metal systems, it is first and foremost a tool for managing virtual systems, and future development will be much more focused on virtual systems than bare-metal systems Virt-Factory provides both a web UI, for ease of use, and an XMLRPC API, for scripting of admin actions. Virt-Factory is built on open-source projects including Cobbler http://cobbler.et.redhat.com, libvirt http://libvirt.org, and Puppet http://reductivelabs.com. Today, Virt-Factory provisions and manages hosts and guests, and addresses some of the problems specific to virtual systems: it creates complete host and guest images from metadata descriptions and centrally manages existing images. Future work will make it possible to abstract away individual hosts and place guests into a pool of equivalent hosts, simplifying the administrator's view of the data center for many tasks. === et-mgmt-tools at redhat.com is the project mailing list... and we already have a git code repository up at the URL above, plus a yum repository, tarballs, and so forth. Questions, comments, ideas? We'd be glad to hear from you. Sincerely, The Virt Factory Team Development: Michael DeHaan, Adrian Likins, Scott Seago Lots of Help From: David Lutterkort, Jim Meyering, and the rest of the Red Hat Emerging Technologies group From mdehaan at redhat.com Thu Mar 15 23:36:53 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 15 Mar 2007 19:36:53 -0400 Subject: [et-mgmt-tools] Cobbler "Roadmap" Message-ID: <45F9D895.9050706@redhat.com> There have been a lot of good requests for cobbler/koan features (and bugfixes) lately, so I thought I would share the current list as I have it. If you have any other requests, now would be a good time to share them. This is the (basically) the same TODO file that's in source control. ==== cobbler TODO list now with 1-5 priority ratings always: 2 - better documentation (particularly online articles) more new thoughts: 1 - re-support repo add taking http:// urls, which will use repotrack/reposync core: 1 - import over http://, ftp:// in addition to existing file and rsync:// and ssh:// 1 - better support for import from DVD (with cleaner imported names) 5 - "check" should check if services (esp httpd) are started 3 - allow rsync without ssh option user interface: 5 - "cobbler add" should be distinguished from "edit" so as to not overwrite 5 - GTK (or Web) UI for Cobbler ? 5 - have -v or --version 3 - cobbler system names (indeed, everything) should be case insensitive. virt: 1 - enable more virtualization types in koan 4 - (more of a koan item) LVM support code cleanup: 5 - remove the cobbler_msg logic since i18n isn't likely ideas: 5 - make the multiple-bootloader system more plugin oriented to allow for alternative bootloaders (not neccessarily kickstart related) to be added. (Low priority) 4 - someone suggested looking at OMAPI for DHCP management small bugs: 3 - cobbler report with a random parameter produces no output that should be an error... is that still the case? 3 - no logrotate or cleanup for /var/log/cobbler (kickstart tracking) 4 - cobbler import should attach "tree" to distro, not profile === One other item not on this list is to move source control over to git... If I had a number #1 thing I want to fix soon it's the DVD import use case. Setting up cobbler using a DVD image (instead of an rsync mirror) deserves to be really really easy, so I will probably tackle that for the next release. --Michael From fj7716hz at aa.jp.fujitsu.com Fri Mar 16 03:03:58 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Fri, 16 Mar 2007 12:03:58 +0900 Subject: [et-mgmt-tools] [PATCH] Check the making domain's disk sources In-Reply-To: <20070315121503.GA1253@redhat.com> References: <45F907A6.6070100@aa.jp.fujitsu.com> <20070315121503.GA1253@redhat.com> Message-ID: <45FA091E.8030507@aa.jp.fujitsu.com> Hi, Dan Daniel P. Berrange wrote: >> The virt-install command specify the Domain's disk sources. >> But the virt-install command doesn't check that the disk sources are >> conflicted among the other domains. >> >> The attached patch resolve this issue in the following way: >> >> 1) Count that the making domian's disk src is already use >> the exists domain's disk source. >> 2) if count > 0, print confirmation message. >> 3) if type yes then leave disk source as it is, >> else ask disk source again. >> >> I am making a patch for virt-manager about this issue. > > If you the non-interactive bit of code into the virtinst/Guest.py module > in the VirtualDisk class, we can share 95% of it between both virtinstall > & virt-manager. Thank you for your suggestion. I rewrite the patch. Signed-off-by: Tatsuro Enokura Thanks, Tatsuro Enokura ----------------------------------------------------------------------------- diff -r ecc4386895aa virt-install --- a/virt-install Thu Mar 15 12:09:51 2007 -0400 +++ b/virt-install Fri Mar 16 11:32:30 2007 +0900 @@ -103,7 +103,7 @@ def get_keymap(keymap, guest): except ValueError, e: print "ERROR: ", e -def get_disk(disk, size, sparse, guest, hvm): +def get_disk(disk, size, sparse, guest, hvm, conn): # FIXME: need to handle a list of disks at some point while 1: msg = "What would you like to use as the disk (path)?" @@ -123,6 +123,23 @@ def get_disk(disk, size, sparse, guest, try: d = virtinst.VirtualDisk(disk, size, sparse = sparse) + if d.is_conflict_disk(conn) is True: + while 1: + retryFlg = False + warnmsg = "Disk %s is already in use by another guest!" % disk + res = prompt_for_input(warnmsg + " Do you really want to use the disk (yes or no)? ") + try: + if yes_or_no(res) is True: + break + else: + retryFlg = True + break + except ValueError, e: + print "ERROR: ", e + continue + if retryFlg is True: + disk = size = None + continue if d.type == virtinst.VirtualDisk.TYPE_FILE and not(hvm) and virtinst.util.is_blktap_capable(): d.driver_name = virtinst.VirtualDisk.DRIVER_TAP except ValueError, e: @@ -133,7 +150,7 @@ def get_disk(disk, size, sparse, guest, guest.disks.append(d) break -def get_disks(disk, size, sparse, guest, hvm): +def get_disks(disk, size, sparse, guest, hvm, conn): # ensure we have equal length lists if (type(disk) == type(size) == list): if len(disk) != len(size): @@ -145,13 +162,13 @@ def get_disks(disk, size, sparse, guest, disk = [ None ] * len(size) if (type(disk) == list): - map(lambda d, s: get_disk(d, s, sparse, guest, hvm), + map(lambda d, s: get_disk(d, s, sparse, guest, hvm, conn), disk, size) elif (type(size) == list): - map(lambda d, s: get_disk(d, s, sparse, guest, hvm), + map(lambda d, s: get_disk(d, s, sparse, guest, hvm, conn), disk, size) else: - get_disk(disk, size, sparse, guest, hvm) + get_disk(disk, size, sparse, guest, hvm, conn) def get_network(mac, bridge, guest): if mac == "RANDOM": @@ -461,7 +478,7 @@ def main(): # set up disks get_disks(options.diskfile, options.disksize, options.sparse, - guest, hvm) + guest, hvm, conn) # set up network information get_networks(options.mac, options.bridge, guest) diff -r ecc4386895aa virtinst/Guest.py --- a/virtinst/Guest.py Thu Mar 15 12:09:51 2007 -0400 +++ b/virtinst/Guest.py Fri Mar 16 11:36:50 2007 +0900 @@ -132,6 +132,42 @@ class VirtualDisk: ret += " \n" ret += " \n" return ret + + def is_conflict_disk(self, conn): + vms = [] + # get working domain's name + ids = conn.listDomainsID(); + for id in ids: + vm = conn.lookupByID(id) + vms.append(vm) + # get defined domain + names = conn.listDefinedDomains() + for name in names: + vm = conn.lookupByName(name) + vms.append(vm) + + count = 0 + for vm in vms: + doc = None + try: + doc = libxml2.parseDoc(vm.XMLDesc(0)) + except: + continue + ctx = doc.xpathNewContext() + try: + try: + count += ctx.xpathEval("count(/domain/devices/disk/source[@dev='%s'])" % self.path) + except: + continue + finally: + if ctx is not None: + ctx.xpathFreeContext() + if doc is not None: + doc.freeDoc() + if count > 0: + return True + else: + return False def __repr__(self): return "%s:%s" %(self.type, self.path) ----------------------------------------------------------------------------- From fj1826dm at aa.jp.fujitsu.com Fri Mar 16 06:10:40 2007 From: fj1826dm at aa.jp.fujitsu.com (Masayuki Sunou) Date: Fri, 16 Mar 2007 15:10:40 +0900 Subject: [et-mgmt-tools] [PATCH] Add the option to guard maximums of the number of virtual CPU by the number of physical CPU Message-ID: <200703161510.ACB82872.KN72GE39@aa.jp.fujitsu.com> Hi In virt-install, the maximum of the number of virtual CPU are not guarded by the number of physical CPU. As a result , the domain slows down when the user set many virtual CPUs by mistake. Therefore, I contribute the patch that adds the option to guard maximums of the number of virtual CPU by the number of physical CPU. Signed-off-by: Masayuki Sunou Thanks, Masayuki Sunou ================================================================================ diff -r ecc4386895aa virt-install --- a/virt-install Thu Mar 15 12:09:51 2007 -0400 +++ b/virt-install Fri Mar 16 14:23:30 2007 +0900 @@ -89,12 +89,24 @@ def get_uuid(uuid, guest): except ValueError, e: print "ERROR: ", e -def get_vcpus(vcpus, guest): - if vcpus: - try: - guest.vcpus = vcpus - except ValueError, e: - print "ERROR: ", e +def get_vcpus(vcpus, check_cpu, guest, conn): + while 1: + vcpus = int(prompt_for_input("How many VCPUs should be attached?",vcpus)) + if check_cpu is not None: + hostinfo = conn.getInfo() + cpu_num = hostinfo[4] * hostinfo[5] * hostinfo[6] * hostinfo[7] + if vcpus > cpu_num: + print "ERROR: Set less or equal the number of physical CPU (%d)." %(cpu_num) + print "" + vcpus = None + continue + if vcpus: + try: + guest.vcpus = vcpus + except ValueError, e: + print "ERROR: ", e + break + def get_keymap(keymap, guest): if keymap: @@ -250,6 +262,8 @@ def parse_args(): help="UUID for the guest; if none is given a random UUID will be generated") parser.add_option("", "--vcpus", type="int", dest="vcpus", help="Number of vcpus to configure for your guest") + parser.add_option("", "--check-cpu", action="store_true", dest="check_cpu", + help="the maximum of physical CPU.") # disk options parser.add_option("-f", "--file", type="string", @@ -456,7 +470,7 @@ def main(): get_name(options.name, guest) get_memory(options.memory, guest) get_uuid(options.uuid, guest) - get_vcpus(options.vcpus, guest) + get_vcpus(options.vcpus, options.check_cpu, guest, conn) get_keymap(options.keymap, guest) # set up disks ================================================================================ From sakaia at jp.fujitsu.com Fri Mar 16 09:09:15 2007 From: sakaia at jp.fujitsu.com (Atsushi SAKAI) Date: Fri, 16 Mar 2007 18:09:15 +0900 Subject: [et-mgmt-tools] Question about Guest disk image size limit Message-ID: <200703160909.l2G99e1J024895@fjmscan502.ms.jp.fujitsu.com> Hi, Guest OS disk Image size is limited to 16GB in case we are using virt-manager. Is this any reason to limit this value? This information is written in following file. virt-manager-0.2.6 virt-manager.glade virt-manager-0.3.1 vmm-create.glade Thanks Atsushi Sakai From fred at x-tend.be Fri Mar 16 09:35:14 2007 From: fred at x-tend.be (=?ISO-8859-1?Q?Fr=E9d=E9ric?= Descamps) Date: Fri, 16 Mar 2007 10:35:14 +0100 Subject: [et-mgmt-tools] Cobbler "Roadmap" In-Reply-To: <45F9D895.9050706@redhat.com> References: <45F9D895.9050706@redhat.com> Message-ID: <1174037714.23852.7.camel@localhost> On Thu, 2007-03-15 at 19:36 -0400, Michael DeHaan wrote: > There have been a lot of good requests for cobbler/koan features (and > bugfixes) lately, so I thought I would share the current list as I have it. > > If you have any other requests, now would be a good time to share > them. This is the (basically) the same TODO file that's in source > control. > > ==== > > cobbler TODO list > now with 1-5 priority ratings > > always: > 2 - better documentation (particularly online articles) > > more new thoughts: > 1 - re-support repo add taking http:// urls, which will use > repotrack/reposync > > core: > 1 - import over http://, ftp:// in addition to existing file and > rsync:// and ssh:// > 1 - better support for import from DVD (with cleaner imported names) > 5 - "check" should check if services (esp httpd) are started > 3 - allow rsync without ssh option > > user interface: > 5 - "cobbler add" should be distinguished from "edit" so as to not > overwrite > 5 - GTK (or Web) UI for Cobbler ? > 5 - have -v or --version > 3 - cobbler system names (indeed, everything) should be case insensitive. > > virt: > 1 - enable more virtualization types in koan > 4 - (more of a koan item) LVM support > > code cleanup: > 5 - remove the cobbler_msg logic since i18n isn't likely > > ideas: > 5 - make the multiple-bootloader system more plugin oriented to allow > for alternative bootloaders (not neccessarily kickstart related) > to be added. (Low priority) > 4 - someone suggested looking at OMAPI for DHCP management > > small bugs: > 3 - cobbler report with a random parameter produces no output > that should be an error... is that still the case? > 3 - no logrotate or cleanup for /var/log/cobbler (kickstart tracking) > 4 - cobbler import should attach "tree" to distro, not profile > > === > > One other item not on this list is to move source control over to git... > > If I had a number #1 thing I want to fix soon it's the DVD import use > case. Setting up cobbler using a DVD image (instead of an rsync > mirror) deserves to be really really easy, so I will probably tackle > that for the next release. > > --Michael > > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > I'd like if possible to add my "request" from Wed, 14 Mar 2007 00:00:45 +0100 into the roadmap too : - the possibility to install on one or more LVM logical volumes instead of xvda. Thank you to take this in consideration. Fred. From fj0873gn at aa.jp.fujitsu.com Fri Mar 16 09:51:03 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Fri, 16 Mar 2007 18:51:03 +0900 Subject: [et-mgmt-tools] [PATCH] RHEL5.0 guest cannot be installed by HTTP and NFS Message-ID: <200703161851.GDF34875.04HGO98K@aa.jp.fujitsu.com> Hi, The RHEL5.0 guest domain cannot be installed by HTTP and NFS. "RPM-GPG-KEY" doesn't exist in RHEL5.0. Though I don't have confidence whether this way is the best, how about this correction? Signed-off-by: Nobuhiro Itou Thanks, Nobuhiro Itou. diff -r ecc4386895aa virtinst/DistroManager.py --- a/virtinst/DistroManager.py Thu Mar 15 12:09:51 2007 -0400 +++ b/virtinst/DistroManager.py Fri Mar 16 17:10:05 2007 +0900 @@ -129,6 +129,8 @@ class MountedImageFetcher(ImageFetcher): file = open(src, "r") except IOError, e: raise RuntimeError, "Invalid location given: " + str(e) + except OSError, (errno, msg): + raise RuntimeError, "Invalid location given: " + msg tmpname = self.saveTemp(file, prefix=base + ".") logging.debug("Saved file to " + tmpname) return tmpname @@ -189,6 +191,14 @@ class FedoraImageStore(ImageStore): return True except RuntimeError, e: logging.debug("Doesn't look like a Fedora distro " + str(e)) + pass + + try: + ignore = fetcher.acquireFile("RPM-GPG-KEY-redhat-release", progresscb) + logging.debug("Detected a RHEL5.x distro") + return True + except RuntimeError, e: + logging.debug("Doesn't look like a RHEL5.x distro " + str(e)) pass finally: if ignore is not None: From berrange at redhat.com Fri Mar 16 12:55:38 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Fri, 16 Mar 2007 12:55:38 +0000 Subject: [et-mgmt-tools] Question about Guest disk image size limit In-Reply-To: <200703160909.l2G99e1J024895@fjmscan502.ms.jp.fujitsu.com> References: <200703160909.l2G99e1J024895@fjmscan502.ms.jp.fujitsu.com> Message-ID: <20070316125537.GA23824@redhat.com> On Fri, Mar 16, 2007 at 06:09:15PM +0900, Atsushi SAKAI wrote: > Hi, > > Guest OS disk Image size is limited to 16GB > in case we are using virt-manager. > Is this any reason to limit this value? No, it seems bogus. It should be possible to create files as large as the underlying filesystem supports (typically 2TB) > This information is written in following file. > virt-manager-0.2.6 virt-manager.glade > virt-manager-0.3.1 vmm-create.glade Dan -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From mdehaan at redhat.com Fri Mar 16 13:03:46 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 16 Mar 2007 09:03:46 -0400 Subject: [et-mgmt-tools] Cobbler "Roadmap" In-Reply-To: <1174037714.23852.7.camel@localhost> References: <45F9D895.9050706@redhat.com> <1174037714.23852.7.camel@localhost> Message-ID: <45FA95B2.8000608@redhat.com> Fr?d?ric Descamps wrote: > On Thu, 2007-03-15 at 19:36 -0400, Michael DeHaan wrote: > >> There have been a lot of good requests for cobbler/koan features (and >> bugfixes) lately, so I thought I would share the current list as I have it. >> >> If you have any other requests, now would be a good time to share >> them. This is the (basically) the same TODO file that's in source >> control. >> >> ==== >> >> cobbler TODO list >> now with 1-5 priority ratings >> >> always: >> 2 - better documentation (particularly online articles) >> >> more new thoughts: >> 1 - re-support repo add taking http:// urls, which will use >> repotrack/reposync >> >> core: >> 1 - import over http://, ftp:// in addition to existing file and >> rsync:// and ssh:// >> 1 - better support for import from DVD (with cleaner imported names) >> 5 - "check" should check if services (esp httpd) are started >> 3 - allow rsync without ssh option >> >> user interface: >> 5 - "cobbler add" should be distinguished from "edit" so as to not >> overwrite >> 5 - GTK (or Web) UI for Cobbler ? >> 5 - have -v or --version >> 3 - cobbler system names (indeed, everything) should be case insensitive. >> >> virt: >> 1 - enable more virtualization types in koan >> 4 - (more of a koan item) LVM support >> >> code cleanup: >> 5 - remove the cobbler_msg logic since i18n isn't likely >> >> ideas: >> 5 - make the multiple-bootloader system more plugin oriented to allow >> for alternative bootloaders (not neccessarily kickstart related) >> to be added. (Low priority) >> 4 - someone suggested looking at OMAPI for DHCP management >> >> small bugs: >> 3 - cobbler report with a random parameter produces no output >> that should be an error... is that still the case? >> 3 - no logrotate or cleanup for /var/log/cobbler (kickstart tracking) >> 4 - cobbler import should attach "tree" to distro, not profile >> >> === >> >> One other item not on this list is to move source control over to git... >> >> If I had a number #1 thing I want to fix soon it's the DVD import use >> case. Setting up cobbler using a DVD image (instead of an rsync >> mirror) deserves to be really really easy, so I will probably tackle >> that for the next release. >> >> --Michael >> >> >> >> _______________________________________________ >> et-mgmt-tools mailing list >> et-mgmt-tools at redhat.com >> https://www.redhat.com/mailman/listinfo/et-mgmt-tools >> >> >> > I'd like if possible to add my "request" from Wed, 14 Mar 2007 00:00:45 > +0100 into the roadmap too : > > - the possibility to install on one or more LVM logical volumes instead > of xvda. > > Thank you to take this in consideration. > > Fred. > > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > That's already there. Also to add: investigate keeping kernel options shorter in some way, (possibly symlink tree URLs?). From hbrock at redhat.com Fri Mar 16 18:04:05 2007 From: hbrock at redhat.com (Hugh Brock) Date: Fri, 16 Mar 2007 14:04:05 -0400 Subject: [et-mgmt-tools] [PATCH] Add the option to guard maximums of the number of virtual CPU by the number of physical CPU In-Reply-To: <200703161510.ACB82872.KN72GE39@aa.jp.fujitsu.com> References: <200703161510.ACB82872.KN72GE39@aa.jp.fujitsu.com> Message-ID: <45FADC15.7090207@redhat.com> Masayuki Sunou wrote: > Hi > > In virt-install, the maximum of the number of virtual CPU are not guarded > by the number of physical CPU. > As a result , the domain slows down when the user set many virtual CPUs > by mistake. > > Therefore, I contribute the patch that adds the option to guard maximums > of the number of virtual CPU by the number of physical CPU. > > > Signed-off-by: Masayuki Sunou > > Thanks, > Masayuki Sunou Actually, we had included this limit early on, but we decided to merely warn the user that excess vcpus can be bad. The thought was that a user might define a domain somewhere, test it, then migrate it to a machine with more CPUs, in which case we would need to allow them to define more virtual CPUs than are physically present. Unless you can think of a compelling reason not to allow a user to do the above, I think we should not take this patch and leave the vcpu setting unbounded. However if you would like to send a patch that warns the user without imposing a hard limit, I think that would be reasonable. Take care, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From hbrock at redhat.com Fri Mar 16 18:34:34 2007 From: hbrock at redhat.com (Hugh Brock) Date: Fri, 16 Mar 2007 14:34:34 -0400 Subject: [et-mgmt-tools] [PATCH] Check the making domain's disk sources In-Reply-To: <45FA091E.8030507@aa.jp.fujitsu.com> References: <45F907A6.6070100@aa.jp.fujitsu.com> <20070315121503.GA1253@redhat.com> <45FA091E.8030507@aa.jp.fujitsu.com> Message-ID: <45FAE33A.3090405@redhat.com> Tatsuro Enokura wrote: > Hi, Dan > > Daniel P. Berrange wrote: >>> The virt-install command specify the Domain's disk sources. >>> But the virt-install command doesn't check that the disk sources are >>> conflicted among the other domains. >>> >>> The attached patch resolve this issue in the following way: >>> >>> 1) Count that the making domian's disk src is already use >>> the exists domain's disk source. >>> 2) if count > 0, print confirmation message. >>> 3) if type yes then leave disk source as it is, >>> else ask disk source again. >>> >>> I am making a patch for virt-manager about this issue. >> If you the non-interactive bit of code into the virtinst/Guest.py module >> in the VirtualDisk class, we can share 95% of it between both virtinstall >> & virt-manager. > > Thank you for your suggestion. > I rewrite the patch. > > Signed-off-by: Tatsuro Enokura > This looks fine, I have committed it. --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From hbrock at redhat.com Fri Mar 16 18:48:00 2007 From: hbrock at redhat.com (Hugh Brock) Date: Fri, 16 Mar 2007 14:48:00 -0400 Subject: [et-mgmt-tools] Question about Guest disk image size limit In-Reply-To: <20070316125537.GA23824@redhat.com> References: <200703160909.l2G99e1J024895@fjmscan502.ms.jp.fujitsu.com> <20070316125537.GA23824@redhat.com> Message-ID: <45FAE660.2010807@redhat.com> Daniel P. Berrange wrote: > On Fri, Mar 16, 2007 at 06:09:15PM +0900, Atsushi SAKAI wrote: >> Hi, >> >> Guest OS disk Image size is limited to 16GB >> in case we are using virt-manager. >> Is this any reason to limit this value? > > No, it seems bogus. It should be possible to create files as large as the > underlying filesystem supports (typically 2TB) > >> This information is written in following file. >> virt-manager-0.2.6 virt-manager.glade >> virt-manager-0.3.1 vmm-create.glade > I have fixed this in the tree. The upper limit on file size is now ~4 TB, which "should be enough for anybody". --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From apuch at redhat.com Fri Mar 16 20:56:04 2007 From: apuch at redhat.com (Andrew Puch) Date: Fri, 16 Mar 2007 16:56:04 -0400 Subject: [et-mgmt-tools] missing profiles ? Message-ID: <1174078564.3972.32.camel@dhcp-64-102-47-235.cisco.com> http://et.redhat.com/page/VF_Install_and_Setup_Guide#Creating_Profiles Hey where is the profiles ? http://virt-factory.et.redhat.com/profiles ? should be http://virt-factory.et.redhat.com/download/profiles/ Also I think this needs to be run before puppetmaster start touch /etc/puppet/manifests/site.pp /usr/sbin/service puppetmaster start From mdehaan at redhat.com Fri Mar 16 22:54:46 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 16 Mar 2007 18:54:46 -0400 Subject: [et-mgmt-tools] missing profiles ? In-Reply-To: <1174078564.3972.32.camel@dhcp-64-102-47-235.cisco.com> References: <1174078564.3972.32.camel@dhcp-64-102-47-235.cisco.com> Message-ID: <45FB2036.7090205@redhat.com> Andrew Puch wrote: > http://et.redhat.com/page/VF_Install_and_Setup_Guide#Creating_Profiles > > Hey where is the profiles ? > > http://virt-factory.et.redhat.com/profiles ? > > should be > http://virt-factory.et.redhat.com/download/profiles/ > > > Also I think this needs to be run before puppetmaster start > > touch /etc/puppet/manifests/site.pp > /usr/sbin/service puppetmaster start > > > > > > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > Yup, I think we've made both corrections by now. Thanks! --Michael From hbrock at redhat.com Sun Mar 18 17:59:09 2007 From: hbrock at redhat.com (Hugh Brock) Date: Sun, 18 Mar 2007 13:59:09 -0400 Subject: [et-mgmt-tools] Re: Virt Manager Problem In-Reply-To: <000501c76986$b8166cc0$28434640$@com> References: <003c01c76919$41937470$c4ba5d50$@com> <45FD7B3A.1070008@redhat.com> <000501c76986$b8166cc0$28434640$@com> Message-ID: <45FD7DED.9060608@redhat.com> FA wrote: > Dear Hugh > Is it possible to port then virt-manager from FC7 to RHEL 5? > > Yours > > Federico Well, you would have to port libvirt, python-virtinst, and virt-manager. We wouldn't support it on RHEL5, but it *might* work. Take care, --Hugh > -----Original Message----- > From: Hugh Brock [mailto:hbrock at redhat.com] > Sent: Sunday, March 18, 2007 1:48 PM > To: FA > Cc: berrange at redhat.com > Subject: Re: Virt Manager Problem > > FA wrote: >> Dear Red Hatters >> >> >> >> I successfully created a new virtual machine using the virtual manager, >> and it rebooted at the end of the process, but it never shows up >> >> in virt-manager. If I try to open it from the file,("restore saved >> machine") it says: >> >> >> >> "Error restoring domain /rhinst/vm1". Is the domain already running" >> >> >> How do I change the Domain for the machine whose file name is at >> /rhinst/vm1 ?? >> > > virt-manager/libvirt in RHEL5 (and currently in FC6) does not have > Inactive Domain Support -- this mean it doesn't know how to tell you > about the existence or state of an inactive domain (like the one you > just created). The only way to take the domain you just created and make > it active is to execute "xm create " as root from the > command line; this will start your domain and it will then show up in > the virt-manager display. > > Note that in Fedora 7 (and soon in FC6) we have added inactive domain > support to libvirt and virt-manager, which means that your domain shows > up in the virt-manager window even though it isn't running. You can then > click on it and start it, or start it from the command line using > libvirt's "virsh" command-line interface. > > Hope this helps, > --Hugh > -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From fj1826dm at aa.jp.fujitsu.com Mon Mar 19 07:38:10 2007 From: fj1826dm at aa.jp.fujitsu.com (Masayuki Sunou) Date: Mon, 19 Mar 2007 16:38:10 +0900 Subject: [et-mgmt-tools] [PATCH] Add the option to guard maximums of thenumber of virtual CPU by the number of physical CPU In-Reply-To: <45FADC15.7090207@redhat.com> References: <200703161510.ACB82872.KN72GE39@aa.jp.fujitsu.com> <45FADC15.7090207@redhat.com> Message-ID: <200703191638.HDH34859.239NEKG7@aa.jp.fujitsu.com> Hi Hugh > Unless you can think of a compelling reason not to allow a user to do > the above, I think we should not take this patch and leave the vcpu > setting unbounded. However if you would like to send a patch that warns > the user without imposing a hard limit, I think that would be reasonable. > I accepted your suggestion, and I made a patch again. This patch adds the option to warn when the number of virtual CPU is more than the number of physical CPU. Signed-off-by: Masayuki Sunou Thanks =============================================================================== diff -r 73d38353c139 virt-install --- a/virt-install Fri Mar 16 14:32:06 2007 -0400 +++ b/virt-install Mon Mar 19 14:44:04 2007 +0900 @@ -89,12 +89,23 @@ def get_uuid(uuid, guest): except ValueError, e: print "ERROR: ", e -def get_vcpus(vcpus, guest): - if vcpus: - try: +def get_vcpus(vcpus, check_cpu, guest, conn): + while 1: + if check_cpu is None: + break + hostinfo = conn.getInfo() + cpu_num = hostinfo[4] * hostinfo[5] * hostinfo[6] * hostinfo[7] + if vcpus <= cpu_num: + break + res = prompt_for_input("Virtual CPU (%d) is more than physical CPU (%d). Would you like to attach virtual CPU more than physical CPU? (yes or no)" %(vcpus, cpu_num)) + try: + if yes_or_no(res): + break + vcpus = int(prompt_for_input("How many VCPUs should be attached?")) + except ValueError, e: + print "ERROR: ", e + if vcpus: guest.vcpus = vcpus - except ValueError, e: - print "ERROR: ", e def get_keymap(keymap, guest): if keymap: @@ -267,6 +278,8 @@ def parse_args(): help="UUID for the guest; if none is given a random UUID will be generated") parser.add_option("", "--vcpus", type="int", dest="vcpus", help="Number of vcpus to configure for your guest") + parser.add_option("", "--check-cpu", action="store_true", dest="check_cpu", + help="Check the maximum of physical CPU.") # disk options parser.add_option("-f", "--file", type="string", @@ -473,7 +486,7 @@ def main(): get_name(options.name, guest) get_memory(options.memory, guest) get_uuid(options.uuid, guest) - get_vcpus(options.vcpus, guest) + get_vcpus(options.vcpus, options.check_cpu, guest, conn) get_keymap(options.keymap, guest) # set up disks =============================================================================== In message <45FADC15.7090207 at redhat.com> "Re: [et-mgmt-tools] [PATCH] Add the option to guard maximums of thenumber of virtual CPU by the number of physical CPU" "Hugh Brock " wrote: > Masayuki Sunou wrote: > > Hi > > > > In virt-install, the maximum of the number of virtual CPU are not guarded > > by the number of physical CPU. > > As a result , the domain slows down when the user set many virtual CPUs > > by mistake. > > > > Therefore, I contribute the patch that adds the option to guard maximums > > of the number of virtual CPU by the number of physical CPU. > > > > > > Signed-off-by: Masayuki Sunou > > > > Thanks, > > Masayuki Sunou > > Actually, we had included this limit early on, but we decided to merely > warn the user that excess vcpus can be bad. The thought was that a user > might define a domain somewhere, test it, then migrate it to a machine > with more CPUs, in which case we would need to allow them to define more > virtual CPUs than are physically present. > > Unless you can think of a compelling reason not to allow a user to do > the above, I think we should not take this patch and leave the vcpu > setting unbounded. However if you would like to send a patch that warns > the user without imposing a hard limit, I think that would be reasonable. > > Take care, > --Hugh > > -- > Red Hat Virtualization Group http://redhat.com/virtualization > Hugh Brock | virt-manager http://virt-manager.org > hbrock at redhat.com | virtualization library http://libvirt.org > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From fj7716hz at aa.jp.fujitsu.com Mon Mar 19 09:08:17 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Mon, 19 Mar 2007 18:08:17 +0900 Subject: [et-mgmt-tools] [PATCH] Check the making domain's mac address(part 2) Message-ID: <45FE5301.3020300@aa.jp.fujitsu.com> Hi, I made the patch the running domain's mac address check(changeset 125), I make the patch in this time that the virt-install comand adds the confirming message. The attached patch the following way: 1) Count that the making domian's MAC address is already use the exists domain's MAC address. 2) Check the hots's NIC MAC address. 2) if count > 0 or , print confirmation message. 3) if type yes then leave disk source as it is, else ask disk source again. Signed-off-by: Tatsuro Enokura Thanks, Tatsuro Enokura -------------------------------------------------------------------------- diff -r 73d38353c139 virt-install --- a/virt-install Fri Mar 16 14:32:06 2007 -0400 +++ b/virt-install Mon Mar 19 16:23:03 2007 +0900 @@ -170,13 +170,64 @@ def get_disks(disk, size, sparse, guest, else: get_disk(disk, size, sparse, guest, hvm, conn) -def get_network(mac, bridge, guest): +def get_network(mac, bridge, gues, conn): if mac == "RANDOM": mac = None n = virtinst.VirtualNetworkInterface(mac, bridge) + if mac is not None: + # get all Domains + ids = conn.listDomainsID(); + vms = [] + for id in ids: + if id == 0: + continue + vm = conn.lookupByID(id) + vms.append(vm) + # get defined domain + names = conn.listDefinedDomains() + for name in names: + vm = conn.lookupByName(name) + vms.append(vm) + # get the Host's NIC MACaddress + hostdevs = virtinst.util.get_host_network_devices() + + while 1: + if mac is None: + msg = "What is the MAC address of your virtual machine?" + mac = prompt_for_input(msg, mac) + n = virtinst.VirtualNetworkInterface(mac, bridge) + + try: + if n.countMACaddr(vms) > 0: + raise ValueError + for (dummy, dummy, dummy, dummy, host_macaddr) in hostdevs: + if mac.upper() == host_macaddr.upper(): + raise ValueError + except ValueError: + while 1: + retryFlg = False + warnmsg = "The MAC address you entered already in use by another guest. " + warnmsg += "The MAC address should be unique in the system. \n" + res = prompt_for_input(warnmsg + " Do you really want to use the MAC address (yes or no)? ") + try: + if yes_or_no(res) is True: + break + else: + retryFlg = True + break + except ValueError, e: + print "ERROR: ", e + continue + if retryFlg is True: + mac = None + continue + else: + break + except: + raise guest.nics.append(n) -def get_networks(macs, bridges, guest): +def get_networks(macs, bridges, guest, conn): # ensure we have equal length lists if (type(macs) == type(bridges) == list): if len(macs) != len(bridges): @@ -188,9 +239,9 @@ def get_networks(macs, bridges, guest): macs = [ None ] * len(bridges) if (type(macs) == list): - map(lambda m, b: get_network(m, b, guest), macs, bridges) - else: - get_network(macs, bridges, guest) + map(lambda m, b: get_network(m, b, guest, conn), macs, bridges) + else: + get_network(macs, bridges, guest, conn) def get_graphics(vnc, vncport, nographics, sdl, keymap, guest): if vnc and nographics: @@ -481,7 +532,7 @@ def main(): guest, hvm, conn) # set up network information - get_networks(options.mac, options.bridge, guest) + get_networks(options.mac, options.bridge, guest, conn) # set up graphics information get_graphics(options.vnc, options.vncport, options.nographics, options.sdl, options.keymap, guest) -------------------------------------------------------------------------- From fj0873gn at aa.jp.fujitsu.com Mon Mar 19 10:01:45 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Mon, 19 Mar 2007 19:01:45 +0900 Subject: [et-mgmt-tools] [PATCH] RuntimeError message is error Message-ID: <200703191901.CAB48982.40O8GHK9@aa.jp.fujitsu.com> Hi The RuntimeError message becomes the error "NameError: global name 'os_type' is not defined". The attached patch adds to solve this problem. Signed-off-by: Nobuhiro Itou Thanks, Nobuhiro Itou. diff -r 73d38353c139 virtinst/FullVirtGuest.py --- a/virtinst/FullVirtGuest.py Fri Mar 16 14:32:06 2007 -0400 +++ b/virtinst/FullVirtGuest.py Mon Mar 19 18:26:07 2007 +0900 @@ -72,7 +72,7 @@ class FullVirtGuest(Guest.XenGuest): if FullVirtGuest.OS_TYPES[self._os_type].has_key(val): self._os_variant = val else: - raise RuntimeError, "OS variant %s does not exist in our dictionary for OS type %s" % (val, os_type) + raise RuntimeError, "OS variant %s does not exist in our dictionary for OS type %s" % (val, self._os_type) os_variant = property(get_os_variant, set_os_variant) def set_os_type_parameters(self, os_type, os_variant): From hbrock at redhat.com Tue Mar 20 00:20:02 2007 From: hbrock at redhat.com (Hugh Brock) Date: Mon, 19 Mar 2007 20:20:02 -0400 Subject: [et-mgmt-tools] virt-manager help is ready Message-ID: <45FF28B2.9000902@redhat.com> Hi all. I've just finished creating the online help for virt-manager. I've also added help menu items and help buttons in most places. I'd love to have any input on things you think I've missed, or especially anything I've gotten wrong. Also, those of you who have expressed an interest in translating it, well, have at it! Enjoy, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From fj0873gn at aa.jp.fujitsu.com Tue Mar 20 09:57:41 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Tue, 20 Mar 2007 18:57:41 +0900 Subject: [et-mgmt-tools] Subject:[PATCH] Windows HVM can't continue to install Message-ID: <200703201857.AEF73948.08KGO94H@aa.jp.fujitsu.com> Hi When I install the Windows HVM domain, it can't continue to install after first restart. It needs a virtual CD-ROM device after reboot too. And the rebooting frequency of Windows is more once than that of RHEL. The attached patch adds to solve this problem. Signed-off-by: Nobuhiro Itou Thanks, Nobuhiro Itou. diff -r 73d38353c139 virt-install --- a/virt-install Fri Mar 16 14:32:06 2007 -0400 +++ b/virt-install Tue Mar 20 11:38:45 2007 +0900 @@ -532,6 +532,17 @@ def main(): struct.unpack("H", buf[0x1fe: 0x200]) == (0xaa55,): # things installed enough that we should be able to restart # the domain + if options.os_type == virtinst.FullVirtGuest.TYPE_WINDOWS: + # continue to Windows installation. + dom = guest.continue_install(conscb,progresscb) + if dom is None: + print "Guest installation failed" + sys.exit(0) + elif dom.info()[0] != libvirt.VIR_DOMAIN_SHUTOFF: + # domain seems to be running + print "Domain installation still in progress. You can reconnect " + print "to the console to complete the installation process." + sys.exit(0) print "Guest installation complete... restarting guest." dom.create() guest.connect_console(conscb) diff -r 73d38353c139 virtinst/Guest.py --- a/virtinst/Guest.py Fri Mar 16 14:32:06 2007 -0400 +++ b/virtinst/Guest.py Mon Mar 19 22:00:33 2007 +0900 @@ -551,9 +551,12 @@ class Guest(object): "networks": self._get_network_xml(install), \ "graphics": self._get_graphics_xml(install) } - def get_config_xml(self, install = True): + def get_config_xml(self, install = True, disk_boot = False): if install: - osblob = self._get_install_xml() + if disk_boot: + osblob = self._get_runtime_xml() + else: + osblob = self._get_install_xml() action = "destroy" else: osblob = self._get_runtime_xml() diff -r 73d38353c139 virtinst/FullVirtGuest.py --- a/virtinst/FullVirtGuest.py Fri Mar 16 14:32:06 2007 -0400 +++ b/virtinst/FullVirtGuest.py Tue Mar 20 13:00:40 2007 +0900 @@ -17,26 +17,56 @@ import Guest import Guest import util import DistroManager +import logging class FullVirtGuest(Guest.XenGuest): - OS_TYPES = { "Linux" : { "Red Hat Enterprise Linux AS 2.1/3" : { "acpi" : True, "apic": True }, \ - "Red Hat Enterprise Linux 4" : { "acpi" : True, "apic": True }, \ - "Red Hat Enterprise Linux 5" : { "acpi" : True, "apic": True }, \ - "Fedora Core 4-6" : { "acpi" : True, "apic": True }, \ - "Suse Linux Enterprise Server" : { "acpi" : True, "apic": True }, \ - "Other Linux 2.6 kernel" : { "acpi" : True, "apic": True } }, \ - "Microsoft Windows" : { "Windows 2000" : { "acpi": False, "apic" : False }, \ - "Windows XP" : { "acpi": True, "apic" : True }, \ - "Windows Server 2003" : { "acpi": True, "apic" : True }, \ - "Windows Vista" : { "acpi": True, "apic" : True } }, \ - "Novell Netware" : { "Netware 4" : { "acpi": True, "apic": True }, \ - "Netware 5" : { "acpi": True, "apic": True }, \ - "Netware 6" : { "acpi": True, "apic": True } }, \ - "Sun Solaris" : { "Solaris 10" : { "acpi": True, "apic": True }, \ - "Solaris 9" : { "acpi": True, "apic": True } }, \ - "Other" : { "MS-DOS" : { "acpi": False, "apic" : False }, \ - "Free BSD" : { "acpi": True, "apic" : True }, \ - "Other" : { "acpi": True, "apic" : True } } } + TYPE_LINUX = "Linux" + TYPE_WINDOWS = "Microsoft Windows" + TYPE_NETWARE = "Novell Netware" + TYPE_SOLARIS = "Sun Solaris" + TYPE_OTHER = "Other" + + VAR_RHEL2_3 = "Red Hat Enterprise Linux AS 2.1/3" + VAR_RHEL4 = "Red Hat Enterprise Linux 4" + VAR_RHEL5 = "Red Hat Enterprise Linux 5" + VAR_FEDORA = "Fedora Core 4-6" + VAR_SLES = "Suse Linux Enterprise Server" + VAR_O_LINUX = "Other Linux 2.6 kernel" + + VAR_WIN2K = "Windows 2000" + VAR_WINXP = "Windows XP" + VAR_WINSV2K3 = "Windows Server 2003" + VAR_WINVISTA = "Windows Vista" + + VAR_NETWARE4 = "Netware 4" + VAR_NETWARE5 = "Netware 5" + VAR_NETWARE6 = "Netware 6" + + VAR_SOLARIS10 = "Solaris 10" + VAR_SOLARIS9 = "Solaris 9" + + VAR_MS_DOS = "MS-DOS" + VAR_FREE_BSD = "Free BSD" + VAR_OTHER = "Other" + + OS_TYPES = { TYPE_LINUX : { VAR_RHEL2_3 : { "acpi" : True, "apic": True }, \ + VAR_RHEL4 : { "acpi" : True, "apic": True }, \ + VAR_RHEL5 : { "acpi" : True, "apic": True }, \ + VAR_FEDORA : { "acpi" : True, "apic": True }, \ + VAR_SLES : { "acpi" : True, "apic": True }, \ + VAR_O_LINUX : { "acpi" : True, "apic": True } }, \ + TYPE_WINDOWS : { VAR_WIN2K : { "acpi": False, "apic" : False }, \ + VAR_WINXP : { "acpi": True, "apic" : True }, \ + VAR_WINSV2K3 : { "acpi": True, "apic" : True }, \ + VAR_WINVISTA : { "acpi": True, "apic" : True } }, \ + TYPE_NETWARE : { VAR_NETWARE4 : { "acpi": True, "apic": True }, \ + VAR_NETWARE5 : { "acpi": True, "apic": True }, \ + VAR_NETWARE6 : { "acpi": True, "apic": True } }, \ + TYPE_SOLARIS : { VAR_SOLARIS10 : { "acpi": True, "apic": True }, \ + VAR_SOLARIS9 : { "acpi": True, "apic": True } }, \ + TYPE_OTHER : { VAR_MS_DOS : { "acpi": False, "apic" : False }, \ + VAR_FREE_BSD : { "acpi": True, "apic" : True }, \ + VAR_OTHER : { "acpi": True, "apic" : True } } } def __init__(self, type=None, arch=None, connection=None, hypervisorURI=None, emulator=None): Guest.Guest.__init__(self, type=type, connection=connection, hypervisorURI=hypervisorURI) @@ -190,3 +220,23 @@ class FullVirtGuest(Guest.XenGuest): self.disks.append(Guest.VirtualDisk(cdrom, device=Guest.VirtualDisk.DEVICE_CDROM, readOnly=True, transient=True)) return tmpfiles + + def continue_install(self, consolecb, meter): + install_xml = self.get_config_xml(disk_boot = True) + logging.debug("Starting guest from '%s'" % ( install_xml )) + meter.start(size=None, text="Starting domain...") + self.domain = self.conn.createLinux(install_xml, 0) + if self.domain is None: + raise RuntimeError, "Unable to start domain for guest, aborting installation!" + meter.end(0) + + self.connect_console(consolecb) + + # ensure there's time for the domain to finish destroying if the + # install has finished or the guest crashed + if consolecb: + time.sleep(1) + + # This should always work, because it'll lookup a config file + # for inactive guest, or get the still running install.. + return self.conn.lookupByName(self.name) From geisj at pagestation.com Tue Mar 20 17:46:36 2007 From: geisj at pagestation.com (Jerry Geis) Date: Tue, 20 Mar 2007 13:46:36 -0400 Subject: [et-mgmt-tools] running virt-manager with existing windows or qemu Message-ID: <46001DFC.5080808@pagestation.com> I am running RHEL beta 5 number 2 with the xen kernel. I have an existing disk that has windows XP on it. I wish to run that XP under virt-manager and xen. I dont see how yet. I was looking at the faq at http://virt-manager.et.redhat.com/faq.html and did not see it. Also, I have an image that I played with under qemu - it worked but I could never get networking to work. I was wanting to try that image under xen and virt-manager and I dont see that either. How do I run that image under virt-manager and xen. When I run the virt-manager it always prompts me to install. I already have existing images or partitions. I am running the xen kernel. Jerry From hbrock at redhat.com Tue Mar 20 18:47:19 2007 From: hbrock at redhat.com (Hugh Brock) Date: Tue, 20 Mar 2007 14:47:19 -0400 Subject: [et-mgmt-tools] running virt-manager with existing windows or qemu In-Reply-To: <46001DFC.5080808@pagestation.com> References: <46001DFC.5080808@pagestation.com> Message-ID: <46002C37.30905@redhat.com> Jerry Geis wrote: > I am running RHEL beta 5 number 2 with the xen kernel. > > I have an existing disk that has windows XP on it. > I wish to run that XP under virt-manager and xen. I dont see how yet. > I was looking at the faq at http://virt-manager.et.redhat.com/faq.html > and did not see it. > > Also, I have an image that I played with under qemu - it worked > but I could never get networking to work. I was wanting to > try that image under xen and virt-manager and I dont see that either. > > How do I run that image under virt-manager and xen. > > When I run the virt-manager it always prompts me to install. > I already have existing images or partitions. > > I am running the xen kernel. > > Jerry Hi Jerry. You can probably get your existing XP install to work under HVM, but to set it up with virt-manager you will actually need to have an XP CD so that the installer has some installation media to look for. In this case you would enter the parameters for your VM and virt-manager would write a config for it and boot your CD. The CD, being Windows, will look at the disk and say "Oh there's an existing install here, I'll boot it." At least this is the way things have been working for me with Windows 2000. A more sensible way to do this is to manually create a config file in /etc/xen that describes your guest -- how much memory, how many vcpus, the address of the guest's disk, and so on. Then start that guest with "xm create ". You should probably make a backup of your XP disk before you do this... Take care, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From jim at rossberry.com Wed Mar 21 00:50:43 2007 From: jim at rossberry.com (Jim Wildman) Date: Tue, 20 Mar 2007 20:50:43 -0400 (EDT) Subject: [et-mgmt-tools] cobbler check and atftpd Message-ID: cobbler check incorrectly reports that #0: need to change field 'server_args' value to '-s /tftpboot' in file '/etc/xinetd.d/tftp' when running with atftp-server-0.7-5.2.el4.rf. This tftp server does not require (or accept) the -s. ------------------------------------------------------------------------ Jim Wildman, CISSP, RHCE jim at rossberry.com http://www.rossberry.com "Society in every state is a blessing, but Government, even in its best state, is a necessary evil; in its worst state, an intolerable one." Thomas Paine From mdehaan at redhat.com Wed Mar 21 14:52:04 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 21 Mar 2007 10:52:04 -0400 Subject: [et-mgmt-tools] cobbler check and atftpd In-Reply-To: References: Message-ID: <46014694.2030507@redhat.com> Jim Wildman wrote: > cobbler check incorrectly reports that > > #0: need to change field 'server_args' value to '-s /tftpboot' in file > '/etc/xinetd.d/tftp' > > when running with atftp-server-0.7-5.2.el4.rf. This tftp server does > not require (or accept) the -s. Cobbler has been tested with tftp-server, so I would recommend ignoring the warning, as it won't affect other cobbler commands. I don't foresee any other problems coming up with atftpd... I'll look into better detection of the tftp-server type in the future. This check is basically there to alert the user if the tftp server location in /var/lib/cobbler/settings (defaults to /tftpboot) does not match what their tftp server is set up to for. If your atftp is configured to use /tftpboot, and that's the value you see in /var/lib/cobbler/settings, you should be good to go. Thanks, Michael > > ------------------------------------------------------------------------ > Jim Wildman, CISSP, RHCE jim at rossberry.com http://www.rossberry.com > "Society in every state is a blessing, but Government, even in its best > state, is a necessary evil; in its worst state, an intolerable one." > Thomas Paine > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From takatom at jp.fujitsu.com Thu Mar 22 09:03:11 2007 From: takatom at jp.fujitsu.com (Takahashi Tomohiro) Date: Thu, 22 Mar 2007 18:03:11 +0900 Subject: [et-mgmt-tools] About a solution to operate the virt-manager in changeset 436 Message-ID: <0e5201c76c60$f0170850$ecb1220a@fj98229700> Hi, I could't operate the virt-manager in changeset 436. when It was connecting Xen hypervisor (Please refer to attachment file :connect.png) note:I'm sorry. Attachment files contain Japanese. It occured following error message. "Unable to open a connection to the Xen hypervisor/daemon." (Please refer to attachment file :error.png and virt-manager.log) I seek a solution to the issue. Please give me an advice, if you know it. (I confirmed that It is able to connect Xen hypervisor in changeset 402) ==virt-manager.log=================== Thur, 22 Mar 2007 15:41:04 ERROR Unable to open connection to hypervisor URI 'xen': virConnect instance has no attribute 'listNetworks' Thur, 22 Mar 2007 15:41:04 ERROR Could not refresh connection xen virConnect instance has no attribute 'listNetworks' ===================================== Thanks, Tomohiro Takahashi. -------------- next part -------------- A non-text attachment was scrubbed... Name: error.png Type: image/png Size: 19755 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: connect.png Type: image/png Size: 18486 bytes Desc: not available URL: From fj7716hz at aa.jp.fujitsu.com Thu Mar 22 11:23:41 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Thu, 22 Mar 2007 20:23:41 +0900 Subject: [et-mgmt-tools] [PATCH] HVM 3rd disk name don't use "hdc" Message-ID: <4602673D.8060802@aa.jp.fujitsu.com> Hi I try that making the HVM domain with 3+ disks by the virt-install command, but the command says following error message: # virt-install --name=HVM_3DISKS --file=/dev/sda6 --file=/dev/sda7 \ --file=/dev/sda8 --hvm --vnc --ram=376 --cdrom=/dev/cdrom libvir: Xen Daemon error : POST operation failed: (xend.err 'Error creating domain: Device hdc (5632, vbd) is already connected.') Traceback (most recent call last): File "/usr/bin/virt-install", line 584, in main() File "/usr/bin/virt-install", line 544, in main dom = guest.start_install(conscb,progresscb) File "/usr/lib/python2.5/site-packages/virtinst/Guest.py", line 626, in start_install return self._do_install(consolecb, meter) File "/usr/lib/python2.5/site-packages/virtinst/Guest.py", line 643, in _do_install self.domain = self.conn.createLinux(install_xml, 0) File "/usr/lib/python2.5/site-packages/libvirt.py", line 429, in createLinux if ret is None:raise libvirtError('virDomainCreateLinux() failed', conn=self) libvirt.libvirtError: virDomainCreateLinux() failed POST operation failed: (xend.err 'Error creating domain: Device hdc (5632, vbd) is already connected.') I think that the 3rd disk's name "hdc" is conflicted CDROM device name. I make the patch resolve this issue. Signed-off-by: Tatsuro Enokura Thanks, Tatsuro Enokura ------------------------------------------------------------ diff -r aec5777422c0 virtinst/Guest.py --- a/virtinst/Guest.py Wed Mar 21 13:46:29 2007 -0400 +++ b/virtinst/Guest.py Thu Mar 22 15:34:06 2007 +0900 @@ -552,8 +552,12 @@ class Guest(object): if d.transient and not install: continue if d.device == VirtualDisk.DEVICE_CDROM and count != 2: - count = 2 - disknode = "%(disknode)s%(dev)c" % { "disknode": self.disknode, "dev": ord('a') + count } + disknode = "%(disknode)s%(dev)c" % { "disknode": self.disknode, "dev": ord('a') + 2 } + elif d.device != VirtualDisk.DEVICE_CDROM and self.disknode == "hd" and count == 2: + count += 1 + disknode = "%(disknode)s%(dev)c" % { "disknode": self.disknode, "dev": ord('a') + count } + else: + disknode = "%(disknode)s%(dev)c" % { "disknode": self.disknode, "dev": ord('a') + count } ret += d.get_xml_config(disknode) count += 1 return ret ------------------------------------------------------------ From Bart.Baars at qnh.nl Thu Mar 22 12:30:19 2007 From: Bart.Baars at qnh.nl (Bart Baars) Date: Thu, 22 Mar 2007 13:30:19 +0100 Subject: [et-mgmt-tools] Cimbiote methods Message-ID: <00042F5A391A1541AA939DEC06CEE7FF10CCCA@EX01.QNH.local> Hi all, For a project at the dutch tax-office, we are planning to use CIM. Our primairy goal is to be able to start/stop services that have a script in /etc/init.d. Unfortunatly, we haven't been able to find a MOF/provider that is able to do this for us. So we decided to write our own. We want to use Cimbiote for this. The example you wrote (example_plugin.py) is pretty clear, except for the methods. I believe there are tho methods, set and get_array_element, and they both require some args. I am trying to access the method using this command: # wbemcli cm 'http://localhost/root/cimv2:ExamplePlugin.Filename="file-9.txt"' get_array_element This returns "a general error occured, not covered by more specific code", and I should check some logs. I have tried various options for the method invocation, but I never get a valid response. Can you give me an example of how to execute this method? Regards, Bart -------------- next part -------------- An HTML attachment was scrubbed... URL: From takatom at jp.fujitsu.com Thu Mar 22 13:01:30 2007 From: takatom at jp.fujitsu.com (Takahashi Tomohiro) Date: Thu, 22 Mar 2007 22:01:30 +0900 Subject: [et-mgmt-tools] About a solution to operate the virt-manager inchangeset 436 References: <0e5201c76c60$f0170850$ecb1220a@fj98229700> Message-ID: <0f6f01c76c82$3b13f360$ecb1220a@fj98229700> Hi, I solved this issue by following fixes. Thanks Tomohiro Takahashi. --- virt-manager--devel.orig/src/virtManager/connection.py 2007-03-22 15:12:00.000000000 +0900 +++ virt-manager--devel/src/virtManager/connection.py 2007-03-22 19:53:56.000000000 +0900 @@ -147,11 +147,12 @@ class vmmConnection(gobject.GObject): def tick(self, noStatsUpdate=False): if self.vmm == None: return - - for name in self.vmm.listNetworks(): - net = self.vmm.networkLookupByName(name) - uuid = self.uuidstr(net.UUID()) - self.nets[uuid] = vmmNetwork(self.config, self, net, uuid) + for n in iter(dir(self.vmm)): + if n == "listNetworks": + for name in self.vmm.listNetworks(): + net = self.vmm.networkLookupByName(name) + uuid = self.uuidstr(net.UUID()) + self.nets[uuid] = vmmNetwork(self.config, self, net, uuid) oldActiveIDs = {} oldInactiveNames = {} > I could't operate the virt-manager in changeset 436. > when It was connecting Xen hypervisor From berrange at redhat.com Thu Mar 22 13:18:31 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 22 Mar 2007 13:18:31 +0000 Subject: [et-mgmt-tools] About a solution to operate the virt-manager in changeset 436 In-Reply-To: <0e5201c76c60$f0170850$ecb1220a@fj98229700> References: <0e5201c76c60$f0170850$ecb1220a@fj98229700> Message-ID: <20070322131831.GB22912@redhat.com> On Thu, Mar 22, 2007 at 06:03:11PM +0900, Takahashi Tomohiro wrote: > Hi, > > I could't operate the virt-manager in changeset 436. > when It was connecting Xen hypervisor > (Please refer to attachment file :connect.png) > note:I'm sorry. Attachment files contain Japanese. > > It occured following error message. > "Unable to open a connection to the Xen hypervisor/daemon." > (Please refer to attachment file :error.png and virt-manager.log) > > I seek a solution to the issue. > Please give me an advice, if you know it. > (I confirmed that It is able to connect Xen hypervisor in changeset 402) > > ==virt-manager.log=================== > Thur, 22 Mar 2007 15:41:04 ERROR Unable to open connection to > hypervisor URI 'xen': > virConnect instance has no attribute 'listNetworks' > Thur, 22 Mar 2007 15:41:04 ERROR Could not refresh connection xen 'exceptions.AttributeError'> virConnect instance > has no attribute 'listNetworks' > ===================================== As of change 436 you need to use libvirt 0.2.1 which adds python APIs for the virNetwork C apis. We use this new networking support in the guest creation wizard. Sorry I forgot to send an announcement / update the README file for this new libvirt requirement. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From mdehaan at redhat.com Thu Mar 22 15:23:44 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 22 Mar 2007 11:23:44 -0400 Subject: [et-mgmt-tools] Cimbiote methods In-Reply-To: <00042F5A391A1541AA939DEC06CEE7FF10CCCA@EX01.QNH.local> References: <00042F5A391A1541AA939DEC06CEE7FF10CCCA@EX01.QNH.local> Message-ID: <46029F80.7030902@redhat.com> Bart Baars wrote: > > Hi all, > > For a project at the dutch tax-office, we are planning to use CIM. Our > primairy goal is to be able to start/stop services that have a script in > /etc/init.d. Unfortunatly, we haven't been able to find a MOF/provider > that > is able to do this for us. So we decided to write our own. > > We want to use Cimbiote for this. The example you wrote > (example_plugin.py) > is pretty clear, except for the methods. I believe there are tho > methods, set and get_array_element, and they both require some args. > > I am trying to access the method using this command: > > # wbemcli cm > 'http://localhost/root/cimv2:ExamplePlugin.Filename="file-9.txt"' > get_array_element > > This returns "a general error occured, not covered by more specific code", > and I should check some logs. I have tried various options for the method > invocation, but I never get a valid response. Can you give me an > example of > how to execute this method? > > Regards, > > Bart > > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools Wow, a post about Cimbiote :) I'm happy and sad at the same time. You're basically the first person who has asked about it. Cimbiote is now largely unmaintained, and I should update the web page to indicate we are looking for a new owner. I would really like to see someone who is well connected with CIM development help continue it. At the time when it was written we saw a lot of large company interest in CIM and wanted to help out in making writing providers easier for those that were not CIM experts -- writing CIM providers is way too painful, and documentation is scarce. Standardized interfaces, we thought, were definitely a good idea, but the problem was these standards were overcomplicated and that created a high barrier to entry for open-source players (who generally preferred simpler and more transparent technology). Unfortunately, we were unable to drum up community interest among pegasus-list and open source CIM project leaders. As we generally prefer secure XMLRPC ourselves (and are also looking forward towards message bus systems like QPid -- http://cwiki.apache.org/qpid/), there wasn't a lot of sense in continuing to invest in it without community interest from a few CIM experts that could help us out. I should probably update the Cimbiote page to better address it's current status. As I recall, the method support did work fine ... though I forget how to use wbemcli myself. There were some strange issues with it and the way it parsed arguments. Most likely wbemcli is not reading the argument list correctly, and we did encounter a few bugs where calling some CIM functions though wbemcli didn't appear to work (using other providers too, not just with Cimbiote). Maybe I can help you out though ... we used some of the sblim providers for reference when working on Cimbiote, and http://sblim.wiki.sourceforge.net/ might have a service provider. I believe it did, though I do not remember exactly what it supported. Hope that helps out some. If you have any interest in contributing to the project, you are definitely welcome to. Thanks, --Michael From berrange at redhat.com Thu Mar 22 15:26:16 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 22 Mar 2007 15:26:16 +0000 Subject: [et-mgmt-tools] [PATCH] HVM 3rd disk name don't use "hdc" In-Reply-To: <4602673D.8060802@aa.jp.fujitsu.com> References: <4602673D.8060802@aa.jp.fujitsu.com> Message-ID: <20070322152616.GD22912@redhat.com> On Thu, Mar 22, 2007 at 08:23:41PM +0900, Tatsuro Enokura wrote: > Hi > > I try that making the HVM domain with 3+ disks by the virt-install command, > but the command says following error message: > # virt-install --name=HVM_3DISKS --file=/dev/sda6 --file=/dev/sda7 \ > --file=/dev/sda8 --hvm --vnc --ram=376 --cdrom=/dev/cdrom This is a ticky one ! We need to a) refuse to allow > 3 --file arguments to be used with HVM guests, and b) make sure the 3rd file gets given hdd as its device name. > diff -r aec5777422c0 virtinst/Guest.py > --- a/virtinst/Guest.py Wed Mar 21 13:46:29 2007 -0400 > +++ b/virtinst/Guest.py Thu Mar 22 15:34:06 2007 +0900 > @@ -552,8 +552,12 @@ class Guest(object): > if d.transient and not install: > continue > if d.device == VirtualDisk.DEVICE_CDROM and count != 2: > - count = 2 > - disknode = "%(disknode)s%(dev)c" % { "disknode": > self.disknode, "dev": ord('a') + count } > + disknode = "%(disknode)s%(dev)c" % { "disknode": > self.disknode, "dev": ord('a') + 2 } > + elif d.device != VirtualDisk.DEVICE_CDROM and self.disknode > == "hd" and count == 2: > + count += 1 > + disknode = "%(disknode)s%(dev)c" % { "disknode": > self.disknode, "dev": ord('a') + count } > + else: > + disknode = "%(disknode)s%(dev)c" % { "disknode": > self.disknode, "dev": ord('a') + count } > ret += d.get_xml_config(disknode) > count += 1 > return ret Your email client has totally mangled the patch, adding in a bunch of extra line breaks :-( Could you re-send it as a plain text attachment instead. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From mdehaan at redhat.com Thu Mar 22 18:35:14 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 22 Mar 2007 14:35:14 -0400 Subject: [et-mgmt-tools] Cimbiote methods In-Reply-To: <46029F80.7030902@redhat.com> References: <00042F5A391A1541AA939DEC06CEE7FF10CCCA@EX01.QNH.local> <46029F80.7030902@redhat.com> Message-ID: <4602CC62.8040008@redhat.com> Michael DeHaan wrote: > Bart Baars wrote: >> >> Hi all, >> >> For a project at the dutch tax-office, we are planning to use CIM. Our >> primairy goal is to be able to start/stop services that have a script in >> /etc/init.d. Unfortunatly, we haven't been able to find a >> MOF/provider that >> is able to do this for us. So we decided to write our own. >> >> We want to use Cimbiote for this. The example you wrote >> (example_plugin.py) >> is pretty clear, except for the methods. I believe there are tho >> methods, set and get_array_element, and they both require some args. >> >> I am trying to access the method using this command: >> >> # wbemcli cm >> 'http://localhost/root/cimv2:ExamplePlugin.Filename="file-9.txt"' >> get_array_element >> >> This returns "a general error occured, not covered by more specific >> code", >> and I should check some logs. I have tried various options for the >> method >> invocation, but I never get a valid response. Can you give me an >> example of >> how to execute this method? >> >> Regards, >> >> Bart >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> et-mgmt-tools mailing list >> et-mgmt-tools at redhat.com >> https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > Wow, a post about Cimbiote :) I'm happy and sad at the same time. > You're basically the first person who has asked about it. Cimbiote > is now largely unmaintained, and I should update the web page to > indicate we are looking for a new owner. I would really like to see > someone who is well connected with CIM development help continue it. > > At the time when it was written we saw a lot of large company interest > in CIM and wanted to help out in making writing providers easier for > those that were not CIM experts -- writing CIM providers is way too > painful, and documentation is scarce. Standardized interfaces, we > thought, were definitely a good idea, but the problem was these > standards were overcomplicated and that created a high barrier to > entry for open-source players (who generally preferred simpler and > more transparent technology). Unfortunately, we were unable to drum > up community interest among pegasus-list and open source CIM project > leaders. As we generally prefer secure XMLRPC ourselves (and are > also looking forward towards message bus systems like QPid -- > http://cwiki.apache.org/qpid/), there wasn't a lot of sense in > continuing to invest in it without community interest from a few CIM > experts that could help us out. I should probably update the > Cimbiote page to better address it's current status. > As I recall, the method support did work fine ... though I forget how > to use wbemcli myself. There were some strange issues with it and > the way it parsed arguments. Most likely wbemcli is not reading the > argument list correctly, and we did encounter a few bugs where calling > some CIM functions though wbemcli didn't appear to work (using other > providers too, not just with Cimbiote). > > Maybe I can help you out though ... we used some of the sblim > providers for reference when working on Cimbiote, and > http://sblim.wiki.sourceforge.net/ might have a service provider. I > believe it did, though I do not remember exactly what it supported. > Hope that helps out some. If you have any interest in contributing > to the project, you are definitely welcome to. > > Thanks, > > --Michael > > > > > > > I just remembered that we were using raw WBEM XML with "wbemexec" in the few instances that wbemcli wasn't working for us. Unfortunately I don't have those xml test files lying around anymore. If you want to get involved with something a bit newer, you might want to look at QPid/AMQP and developing your own way of sending commands to and from your nodes rather than CIM. Message buses and CIM are of course totally different, but I think there is a lower barrier to entry and a lot more flexibility there. In particular, I'm not particularly fond of the way CIM handles events (indications)... --Michael From rjones at redhat.com Thu Mar 22 19:05:31 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Thu, 22 Mar 2007 19:05:31 +0000 Subject: [et-mgmt-tools] [PATCH] For connection errors, show full details Message-ID: <4602D37B.9000509@redhat.com> Currently virt-manager hides the real details behind a connection failure in a log file, which users have to "just know" how to find. This patch adds a new dialog class, vmmErrorDialog, derived from gtk.MessageDialog, which has a "Show details" button to show the true reason for a failure. It modifies the exception handler for connections to give the underlying reason and a stack trace. Also attached, two screenshots so you can see what I mean without needing to apply the patch. Rich. -- Emerging Technologies, Red Hat http://et.redhat.com/~rjones/ 64 Baker Street, London, W1U 7DF Mobile: +44 7866 314 421 "[Negative numbers] darken the very whole doctrines of the equations and make dark of the things which are in their nature excessively obvious and simple" (Francis Maseres FRS, mathematician, 1759) -------------- next part -------------- A non-text attachment was scrubbed... Name: virt-manager-show-details-of-failed-connection.patch Type: text/x-patch Size: 6635 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: connection-failure-1.png Type: image/png Size: 17384 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: connection-failure-2.png Type: image/png Size: 36780 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 3237 bytes Desc: S/MIME Cryptographic Signature URL: From hbrock at redhat.com Thu Mar 22 19:43:16 2007 From: hbrock at redhat.com (Hugh Brock) Date: Thu, 22 Mar 2007 15:43:16 -0400 Subject: [et-mgmt-tools] Subject:[PATCH] Windows HVM can't continue to install In-Reply-To: <200703201857.AEF73948.08KGO94H@aa.jp.fujitsu.com> References: <200703201857.AEF73948.08KGO94H@aa.jp.fujitsu.com> Message-ID: <4602DC54.7050203@redhat.com> Nobuhiro Itou wrote: > Hi > > When I install the Windows HVM domain, it can't continue to install > after first restart. > It needs a virtual CD-ROM device after reboot too. > And the rebooting frequency of Windows is more once than that of RHEL. > The attached patch adds to solve this problem. > > Signed-off-by: Nobuhiro Itou > > > Thanks, > Nobuhiro Itou. > We don't really want to put OS-specific stuff into the virt-install code if we can help it; we made an exception for the config flag setting stuff because we're just abstracting behavior (setting acpi= and apic=) that is already available. However I don't see any reason why a patch that would allow a virtinst argument like "--keep-cdrom" and then put something like "... etc" in the domain's permanent xml wouldn't work. This would mean that the domain would have access to the cdrom drive every time it reboots. Of course, it also means we'll need to provide some way to disconnect the domain from that drive once the install process is finished, but we'll get to that later. Let me know if that works for you... --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From bart.lists at whiteley.org Thu Mar 22 20:53:40 2007 From: bart.lists at whiteley.org (Bart Whiteley) Date: Thu, 22 Mar 2007 14:53:40 -0600 Subject: [et-mgmt-tools] Cimbiote methods In-Reply-To: <46029F80.7030902@redhat.com> References: <00042F5A391A1541AA939DEC06CEE7FF10CCCA@EX01.QNH.local> <46029F80.7030902@redhat.com> Message-ID: On 3/22/07, Michael DeHaan wrote: > > Wow, a post about Cimbiote :) I'm happy and sad at the same time. > You're basically the first person who has asked about it. Cimbiote is > now largely unmaintained, and I should update the web page to indicate > we are looking for a new owner. I would really like to see someone > who is well connected with CIM development help continue it. I would like to take ownership of Cimbiote if you're looking for a new owner. I added support for associations to Cimbiote a couple of months back, but haven't got around to testing it thoroughly yet, which is why I haven't submitted the patch yet. At the time when it was written we saw a lot of large company interest > in CIM and wanted to help out in making writing providers easier for > those that were not CIM experts -- writing CIM providers is way too > painful, and documentation is scarce. Standardized interfaces, we > thought, were definitely a good idea, but the problem was these > standards were overcomplicated and that created a high barrier to entry > for open-source players (who generally preferred simpler and more > transparent technology). Unfortunately, we were unable to drum up > community interest among pegasus-list and open source CIM project > leaders. I did a presentation on pywbem and Cimbiote at Management Developers Conference (mandevcon.com) back in December. I believe the current CIM/WBEM community isn't very interested in a Python provider interface because to date, most of this group has been involved in embedded environments such as BMCs, where python is not guaranteed to be present. -------------- next part -------------- An HTML attachment was scrubbed... URL: From mdehaan at redhat.com Thu Mar 22 21:21:17 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 22 Mar 2007 17:21:17 -0400 Subject: [et-mgmt-tools] Cimbiote methods In-Reply-To: References: <00042F5A391A1541AA939DEC06CEE7FF10CCCA@EX01.QNH.local> <46029F80.7030902@redhat.com> Message-ID: <4602F34D.9090604@redhat.com> Bart Whiteley wrote: > On 3/22/07, *Michael DeHaan* > wrote: > > Wow, a post about Cimbiote :) I'm happy and sad at the same time. > You're basically the first person who has asked about > it. Cimbiote is > now largely unmaintained, and I should update the web page to > indicate > we are looking for a new owner. I would really like to see someone > who is well connected with CIM development help continue it. > > > I would like to take ownership of Cimbiote if you're looking for a new > owner. I added support for associations to Cimbiote a couple of months > back, but haven't got around to testing it thoroughly yet, which is why > I haven't submitted the patch yet. This sounds really good to me. Cimbiote is GPL'd with the plugin examples being LGPL'd, so I do not see a problem with this. Obviously leave copyrights intact :) I would say go ahead and find a new home to host it. > > At the time when it was written we saw a lot of large company interest > in CIM and wanted to help out in making writing providers easier for > those that were not CIM experts -- writing CIM providers is way too > painful, and documentation is scarce. Standardized interfaces, we > thought, were definitely a good idea, but the problem was these > standards were overcomplicated and that created a high barrier to > entry > for open-source players (who generally preferred simpler and more > transparent technology). Unfortunately, we were unable to drum up > community interest among pegasus-list and open source CIM project > leaders. > > > I did a presentation on pywbem and Cimbiote at Management Developers > Conference ( mandevcon.com ) back in December. Very interesting. Have slides? :) > I believe the current > CIM/WBEM community isn't very interested in a Python provider > interface because to date, most of this group has been involved in > embedded environments such as BMCs, where python is not guaranteed > to be present. --Michael > > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From rjones at redhat.com Thu Mar 22 21:31:44 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Thu, 22 Mar 2007 21:31:44 +0000 Subject: [et-mgmt-tools] [PATCH] For connection errors, show full details (version 2) Message-ID: <4602F5C0.9000107@redhat.com> Following on from https://www.redhat.com/archives/et-mgmt-tools/2007-March/msg00174.html, this is a second version of the patch which more or less follows the design of error dialogs used by 'pup'. I converted pup's glade-based configuration file to direct code (the direct code is much shorter for a start). Two screenshots show what it looks like. Rich. -- Emerging Technologies, Red Hat http://et.redhat.com/~rjones/ 64 Baker Street, London, W1U 7DF Mobile: +44 7866 314 421 "[Negative numbers] darken the very whole doctrines of the equations and make dark of the things which are in their nature excessively obvious and simple" (Francis Maseres FRS, mathematician, 1759) -------------- next part -------------- A non-text attachment was scrubbed... Name: Screenshot.png Type: image/png Size: 16916 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: Screenshot-1.png Type: image/png Size: 36939 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: virt-manager-show-details-of-failed-connection-2.patch Type: text/x-patch Size: 5724 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 3237 bytes Desc: S/MIME Cryptographic Signature URL: From berrange at redhat.com Thu Mar 22 23:28:09 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 22 Mar 2007 23:28:09 +0000 Subject: [et-mgmt-tools] [PATCH] For connection errors, show full details (version 2) In-Reply-To: <4602F5C0.9000107@redhat.com> References: <4602F5C0.9000107@redhat.com> Message-ID: <20070322232809.GC1545@redhat.com> On Thu, Mar 22, 2007 at 09:31:44PM +0000, Richard W.M. Jones wrote: > Following on from > https://www.redhat.com/archives/et-mgmt-tools/2007-March/msg00174.html, > this is a second version of the patch which more or less follows the > design of error dialogs used by 'pup'. I converted pup's glade-based > configuration file to direct code (the direct code is much shorter for a > start). Looks good to me. I'll commit it shortly. The last stage of the create wizard could also probably do with a similar adaptation to allow viewing of detailed errors. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From takatom at jp.fujitsu.com Fri Mar 23 05:40:13 2007 From: takatom at jp.fujitsu.com (Takahashi Tomohiro) Date: Fri, 23 Mar 2007 14:40:13 +0900 Subject: [et-mgmt-tools] About a solution to operate the virt-manager inchangeset 436 Message-ID: <019501c76d0d$c898d490$ecb1220a@fj98229700> Hi, Thank you for your comments. It occors the similar error?virt-manager.log), When I used libvirt 0.2.1 ChangeLog 1.482 (Dialog is the same message. Please refer to attachment file connection_error.png note:I'm sorry. Attachment files contain Japanese.) Do you know anything about this error ? ==== virt-manager.log ============================================= Fri, 23 Mar 2007 11:50:40 ERROR Unable to open connection to hypervisor URI 'xen': 'NoneType' object is not iterable Fri, 23 Mar 2007 11:50:41 ERROR Could not refresh connection xen 'NoneType' object is not iterable Fri, 23 Mar 2007 11:50:42 ERROR Could not refresh connection xen 'NoneType' object is not iterable > As of change 436 you need to use libvirt 0.2.1 which adds python APIs for > the virNetwork C apis. We use this new networking support in the guest > creation wizard. Sorry I forgot to send an announcement / update the > README > file for this new libvirt requirement. Thanks, Tomohiro Takahashi. -------------- next part -------------- A non-text attachment was scrubbed... Name: connection_error.png Type: image/png Size: 19755 bytes Desc: not available URL: From cobbler at msquared.id.au Fri Mar 23 06:57:48 2007 From: cobbler at msquared.id.au (Msquared) Date: Fri, 23 Mar 2007 15:57:48 +0900 Subject: [et-mgmt-tools] kickstart file specified in distro (rather than profile)? In-Reply-To: <45E845A0.8090207@redhat.com> References: <20070302071603.GO13894@sliderule.msquared.com.au> <45E845A0.8090207@redhat.com> Message-ID: <20070323065748.GA22919@sliderule.msquared.com.au> On Fri, Mar 02, 2007 at 10:41:20AM -0500, Michael DeHaan wrote: > >Can I specify a kickstart entry in the distro? This would be the default, > >which I could override for each profile I create. > It belongs in the profile. A kickstart explains /how/ a distribution > should be installed, not /what/ distribution should be installed. I don't disagree, but it occurs to me that the kickstart can also explain what /parts of/ a distribution should be installed (via %packages), so in part it is explaining /what/ (albeit not what /distribution/). I still think that you should be able to specify a default kickstart file to associate with the distro (since in a given installation it may be likely that /how/ is the same for all of the configurations, and you only want to customers /what parts/ are installed, via the profile). Or perhaps an ability to build a profile on top of another profile and just customise some of it? Or another layer above or below profiles? Perhaps 'customisation' or something? > >I ask because I notice the kickstart option to cobbler profile add is > >optional. So what kickstart is used if the option is not given explicitly in 'cobbler profile add'? Regards, Msquared... From fj1826dm at aa.jp.fujitsu.com Fri Mar 23 08:37:41 2007 From: fj1826dm at aa.jp.fujitsu.com (Masayuki Sunou) Date: Fri, 23 Mar 2007 17:37:41 +0900 Subject: [et-mgmt-tools] [PATCH][RESEND] Add the option to guard maximums ofthenumber of virtual CPU by the number of physical CPU In-Reply-To: <200703191638.HDH34859.239NEKG7@aa.jp.fujitsu.com> References: <200703161510.ACB82872.KN72GE39@aa.jp.fujitsu.com> <45FADC15.7090207@redhat.com> <200703191638.HDH34859.239NEKG7@aa.jp.fujitsu.com> Message-ID: <200703231737.AAB48410.9GN27EK3@aa.jp.fujitsu.com> Hi, Would you give me a comment on this patch? If not, please apply it. > > Unless you can think of a compelling reason not to allow a user to do > > the above, I think we should not take this patch and leave the vcpu > > setting unbounded. However if you would like to send a patch that warns > > the user without imposing a hard limit, I think that would be reasonable. > > > I accepted your suggestion, and I made a patch again. > > This patch adds the option to warn when the number of virtual CPU is > more than the number of physical CPU. Signed-off-by: Masayuki Sunou Thanks, Masayuki Sunou =============================================================================== diff -r aec5777422c0 virt-install --- a/virt-install Wed Mar 21 13:46:29 2007 -0400 +++ b/virt-install Fri Mar 23 17:16:26 2007 +0900 @@ -89,12 +89,23 @@ def get_uuid(uuid, guest): except ValueError, e: print "ERROR: ", e -def get_vcpus(vcpus, guest): - if vcpus: - try: +def get_vcpus(vcpus, check_cpu, guest, conn): + while 1: + if check_cpu is None: + break + hostinfo = conn.getInfo() + cpu_num = hostinfo[4] * hostinfo[5] * hostinfo[6] * hostinfo[7] + if vcpus <= cpu_num: + break + res = prompt_for_input("Virtual CPU (%d) is more than physical CPU (%d). Would you like to attach virtual CPU more than physical CPU? (yes or no)" %(vcpus, cpu_num)) + try: + if yes_or_no(res): + break + vcpus = int(prompt_for_input("How many VCPUs should be attached?")) + except ValueError, e: + print "ERROR: ", e + if vcpus: guest.vcpus = vcpus - except ValueError, e: - print "ERROR: ", e def get_keymap(keymap, guest): if keymap: @@ -296,6 +307,8 @@ def parse_args(): help="UUID for the guest; if none is given a random UUID will be generated") parser.add_option("", "--vcpus", type="int", dest="vcpus", help="Number of vcpus to configure for your guest") + parser.add_option("", "--check-cpu", action="store_true", dest="check_cpu", + help="Check the maximum of physical CPU.") # disk options parser.add_option("-f", "--file", type="string", @@ -505,7 +518,7 @@ def main(): get_name(options.name, guest) get_memory(options.memory, guest) get_uuid(options.uuid, guest) - get_vcpus(options.vcpus, guest) + get_vcpus(options.vcpus, options.check_cpu, guest, conn) get_keymap(options.keymap, guest) # set up disks =============================================================================== From jeevanullas at gmail.com Fri Mar 23 09:37:00 2007 From: jeevanullas at gmail.com (Deependra Shekhawat) Date: Fri, 23 Mar 2007 05:37:00 -0400 Subject: [et-mgmt-tools] Deependra has Tagged you! :) Message-ID: <200703230937.l2N9b0om022990@mx2.redhat.com> An HTML attachment was scrubbed... URL: From rjones at redhat.com Fri Mar 23 11:03:43 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Fri, 23 Mar 2007 11:03:43 +0000 Subject: [et-mgmt-tools] About a solution to operate the virt-manager inchangeset 436 In-Reply-To: <019501c76d0d$c898d490$ecb1220a@fj98229700> References: <019501c76d0d$c898d490$ecb1220a@fj98229700> Message-ID: <4603B40F.1070005@redhat.com> Takahashi Tomohiro wrote: > Hi, > > Thank you for your comments. > > It occors the similar error?virt-manager.log), When I used libvirt > 0.2.1 ChangeLog 1.482 > (Dialog is the same message. > Please refer to attachment file connection_error.png > note:I'm sorry. Attachment files contain Japanese.) > > Do you know anything about this error ? > > ==== virt-manager.log ============================================= I'm getting a very similar error at the moment, which I'm looking into. Rich. -- Emerging Technologies, Red Hat http://et.redhat.com/~rjones/ 64 Baker Street, London, W1U 7DF Mobile: +44 7866 314 421 "[Negative numbers] darken the very whole doctrines of the equations and make dark of the things which are in their nature excessively obvious and simple" (Francis Maseres FRS, mathematician, 1759) -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 3237 bytes Desc: S/MIME Cryptographic Signature URL: From pasik at iki.fi Fri Mar 23 13:35:06 2007 From: pasik at iki.fi (Pasi =?iso-8859-1?Q?K=E4rkk=E4inen?=) Date: Fri, 23 Mar 2007 15:35:06 +0200 Subject: [et-mgmt-tools] Default Cobbler kickstart_fc6.ks and kernel command line options length Message-ID: <20070323133506.GM29393@edu.joroinen.fi> Hello list! I just set up cobbler, and it seems to work fine with ks.cfg from manually installed system. Then I tried to use the default cobbler fc6 ks.cfg (/etc/cobbler/kickstart_fc6.ks) but I'm a bit confused about some things.. # Use network installation url --url=$tree # If any cobbler repo definitions were referenced in the kickstart profile, include them here. $yum_repo_stanza Those two.. FC6 installation fails and gives errors about those. What did I do wrong? I should have specified some parameters when adding the profile.. ? Then there's another problem.. if I want to give "dd=http://some.server.tld/dd.iso" and "updates=http://some.server.tld/updates.img" as a parameters for the installer, the kernel command line will get too long, and the parameters get cut and then obviously don't work anymore.. Anything I can do about this? Thanks! -- Pasi From mdehaan at redhat.com Fri Mar 23 14:10:48 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 23 Mar 2007 10:10:48 -0400 Subject: [et-mgmt-tools] kickstart file specified in distro (rather than profile)? In-Reply-To: <20070323065748.GA22919@sliderule.msquared.com.au> References: <20070302071603.GO13894@sliderule.msquared.com.au> <45E845A0.8090207@redhat.com> <20070323065748.GA22919@sliderule.msquared.com.au> Message-ID: <4603DFE8.3050803@redhat.com> Msquared wrote: > On Fri, Mar 02, 2007 at 10:41:20AM -0500, Michael DeHaan wrote: > > >>> Can I specify a kickstart entry in the distro? This would be the default, >>> which I could override for each profile I create. >>> > > >> It belongs in the profile. A kickstart explains /how/ a distribution >> should be installed, not /what/ distribution should be installed. >> > > I don't disagree, but it occurs to me that the kickstart can also explain > what /parts of/ a distribution should be installed (via %packages), so in > part it is explaining /what/ (albeit not what /distribution/). > You seem to be missing the design point of the abstraction layer. A given distribution can be installed multiple ways, and folks should be thinking about a role of a system as opposed to just what distribution is assigned to the system. Adding a profile is very easy: cobbler profile add --name=foo --distro=bar --kickstart=/opt/foo/whatever.ks > I still think that you should be able to specify a default kickstart file > to associate with the distro (since in a given installation it may be > likely that /how/ is the same for all of the configurations, and you only > want to customers /what parts/ are installed, via the profile). > If you use the import feature, a profile is created with the same name as the distro automagically with (unless /etc/cobbler/*.ks is modified) an @base kickstart. You can use that. > Or perhaps an ability to build a profile on top of another profile and > just customise some of it? Or another layer above or below profiles? > Perhaps 'customisation' or something? > > Something like this is possible. >>> I ask because I notice the kickstart option to cobbler profile add is >>> optional. >>> > > So what kickstart is used if the option is not given explicitly in > 'cobbler profile add'? > > An interactive install. > Regards, Msquared... > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From hbrock at redhat.com Fri Mar 23 14:12:26 2007 From: hbrock at redhat.com (Hugh Brock) Date: Fri, 23 Mar 2007 10:12:26 -0400 Subject: [et-mgmt-tools] [PATCH][RESEND] Add the option to guard maximums ofthenumber of virtual CPU by the number of physical CPU In-Reply-To: <200703231737.AAB48410.9GN27EK3@aa.jp.fujitsu.com> References: <200703161510.ACB82872.KN72GE39@aa.jp.fujitsu.com> <45FADC15.7090207@redhat.com> <200703191638.HDH34859.239NEKG7@aa.jp.fujitsu.com> <200703231737.AAB48410.9GN27EK3@aa.jp.fujitsu.com> Message-ID: <4603E04A.6090808@redhat.com> Masayuki Sunou wrote: > Hi, > > Would you give me a comment on this patch? > If not, please apply it. > >>> Unless you can think of a compelling reason not to allow a user to do >>> the above, I think we should not take this patch and leave the vcpu >>> setting unbounded. However if you would like to send a patch that warns >>> the user without imposing a hard limit, I think that would be reasonable. >>> >> I accepted your suggestion, and I made a patch again. >> >> This patch adds the option to warn when the number of virtual CPU is >> more than the number of physical CPU. > > Signed-off-by: Masayuki Sunou > > Thanks, > Masayuki Sunou > > Sorry for the delay, I have applied it now. Thanks! --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From mdehaan at redhat.com Fri Mar 23 14:19:55 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 23 Mar 2007 10:19:55 -0400 Subject: [et-mgmt-tools] Default Cobbler kickstart_fc6.ks and kernel command line options length In-Reply-To: <20070323133506.GM29393@edu.joroinen.fi> References: <20070323133506.GM29393@edu.joroinen.fi> Message-ID: <4603E20B.40805@redhat.com> Pasi K?rkk?inen wrote: > Hello list! > > I just set up cobbler, and it seems to work fine with ks.cfg from manually installed > system. > > Then I tried to use the default cobbler fc6 ks.cfg (/etc/cobbler/kickstart_fc6.ks) > but I'm a bit confused about some things.. > This kickstart in /etc/cobbler is used by the "cobbler import" command and is not really intended for use with "profile add". The import command fills in the template variable for "tree" automatically and makes it point to a http:// accessible URL for the kickstart tree. I'm betting you did "cobbler profile add" manually and thus this variable isn't filled in. > # Use network installation > url --url=$tree > > # If any cobbler repo definitions were referenced in the kickstart profile, include them here. > $yum_repo_stanza > > Those two.. FC6 installation fails and gives errors about those. > > I think it most likely gave you an error about the first bit. The second bit is templated out automagically by every "cobbler sync", and I doubt very much it would be wrong. Look in /var/www/cobbler/kickstarts and /var/www/cobbler/kickstarts_sys and you will see how these kickstarts are templated out in reality. If you aren't using "cobbler import" and have added a distro manually, you need to supply your own (different) kickstart to --kickstart that has a non-templated "url" line or otherwise also need to pass in --ksmeta="tree=http://server/path/to/install/source" to the "distro add" command so that the $tree variable can be templated out correctly. > What did I do wrong? I should have specified some parameters when adding the > profile.. ? > > Yes :) > Then there's another problem.. if I want to give > "dd=http://some.server.tld/dd.iso" and "updates=http://some.server.tld/updates.img" > as a parameters for the installer, the kernel command line will get too > long, and the parameters get cut and then obviously don't work anymore.. > Anything I can do about this? > See "cobbler repo add" in the manual for the way to do this. Basically you tell cobbler to mirror repositories like "fc6i386updates" and then mention the repos you want to attach using the "--repos" argument to "cobbler profile add". Cobbler will template out the kickstarts to be able to use those repos during the install process. As for kernel argument lengths, I've done several things that will yield substantially reduced kernel command line lengths in 0.4.4, which I plan to release today after a bit more testing. I'm going to be opening up some Wiki pages soon on et.redhat.com that should serve as a good place to save a lot of cobbler examples and answers to questions like this. > > Thanks! > > > -- Pasi > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From dlutter at redhat.com Fri Mar 23 16:18:49 2007 From: dlutter at redhat.com (David Lutterkort) Date: Fri, 23 Mar 2007 09:18:49 -0700 Subject: [et-mgmt-tools] cft 0.2.0 Message-ID: <1174666729.3528.2.camel@localhost.localdomain> I am pleased to announce the release of cft-0.2.0; the major new feature in this release is that cft now tracks package operations, so that things like yum -y install httpd rpm -e evolution will be reflected in the generated manifest. Details and downloads can be found at http://cft.et.redhat.com/ RPM's for Fedora and RHEL are available at http://people.redhat.com/dlutter/yum David From mdehaan at redhat.com Fri Mar 23 16:59:05 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 23 Mar 2007 12:59:05 -0400 Subject: [et-mgmt-tools] koan on RHEL-4, missing dependencies... Message-ID: <46040759.6030203@redhat.com> I've just noticed that koan on RHEL-4 tries to overzealously import virtualization features, even when not using the --virt-parameter. This can result in --replace-self not working on older systems. If you are trying to get --replace-self in koan to work on system and running into trouble because of missing virtualization modules, here's a patch: diff -r eaeec03f11d1 koan/app.py --- a/koan/app.py Wed Mar 14 16:41:44 2007 -0400 +++ b/koan/app.py Fri Mar 23 12:55:51 2007 -0400 @@ -27,7 +27,6 @@ import shutil import shutil import errno import re -import virtcreate """ koan --virt [--profile=webserver|--system=name] --server=hostname @@ -98,9 +97,6 @@ def main(): except InfoException, ie: print str(ie) return 1 - except virtcreate.VirtCreateException, xce: - print str(xce) - return 2 except: traceback.print_exc() return 3 @@ -587,6 +583,11 @@ class Koan: # parser issues? lang needs a trailing = and somehow doesn't have it. kextra = kextra.replace("lang ","lang= ") + try: + import virtcreate + except: + print "no virtualization support available, install python-virtinst?" + results = virtcreate.start_paravirt_install( name=self.calc_virt_name(pd), ram=self.calc_virt_ram(pd), This will be included in the next release of koan and will checked in shortly. --Michael From mdehaan at redhat.com Fri Mar 23 17:16:50 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 23 Mar 2007 13:16:50 -0400 Subject: [et-mgmt-tools] Cobbler 0.4.4 released Message-ID: <46040B82.6030505@redhat.com> The purpose of this release was to address recently-requested improvements. The main features you will want to be aware of are: (A) the PXE files are now completely templated by various templates stored in /var/www/cobbler. This should allow for user modification of timeouts and other variables (if desirable) as well as adding on additional parameters like the previously requested support for "ipappend 2". (B) much improved import features, which a focus being on the DVD import use case. DVD imports (either mounted regularly or by loopback) are now first class citzens to "cobbler import". Version detection and kickstart assignment is done by looking at versions of the "-release" RPMs. With this modification, starting with a DVD or ISO is now the fastest way to bring up provisioning infrastructure from scratch. Those importing DVD's by using the "ssh hack" don't need to do that anymore, as all imports now mirror automagically. (This may later result in a feature request to do an import without mirroring content, though in most use cases, mirroring is important for accessibility of the files). Also note that "tree" is now a metadata parameter attached to the distro, not the profile (yay!). Also, in bugfix land, this addresses the much reported problem of "my kernel options are way too long". DVD import now names imported distros very cleanly, so the names are a lot shorter. In addition, cobbler makes use of a "links" directory of path symlinks on import, to further shorten the paths. Also note that cobbler has new values for /var/lib/cobbler/settings, removing a few kernel parameters that really don't need to be in the default list. If you have an existing cobbler install, these won't be updated, so you may want to manually remove "ramdisk_size" and "append" from /var/lib/cobbler/settings and re run "cobbler sync". Those parameters don't need to be there. Basically all you really want are "ksdevice" (maybe), "lang", and "text". And maybe not even those depending on what you want to put in your kickstarts. From the CHANGELOG: - Generate PXE configuration files from templates in /etc/cobbler to be more customizable - Fix bug with wrong kickstart metadata being used for import - Fix bug with argument parsing for --repos - Much cleaner distro/profile names with --import - For import, the "tree" parameter is now attached to the distro, not the profile - Add "links" directory in webdir for symlinking to full kickstart tree paths. - Misc tweaks to shorten kernel parameter length - Giving invalid arguments to "report" will show an error message - Distros, Profiles, and System names are now case insensitive. RPMs and code out now, pushing to Fedora's yum shortly... --Michael From mdehaan at redhat.com Fri Mar 23 18:45:43 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 23 Mar 2007 14:45:43 -0400 Subject: [et-mgmt-tools] Bootstrapping cobbler from a DVD image Message-ID: <46042057.2080308@redhat.com> I just wrote this up... http://et.redhat.com/page/Cobbler_Import This basically shows how you can go from "blank system" with no cobbler installed to network boot server really really quickly without having to understand a lot of the details of cobbler. Basically you can set up a boot server, from scratch, in about 20 minutes or so -- and that includes waiting for the files to copy around. More articles/howtos to come over time, I'm sure. --Michael From rnijskens at gmail.com Fri Mar 23 19:05:03 2007 From: rnijskens at gmail.com (Raf Nijskens) Date: Fri, 23 Mar 2007 20:05:03 +0100 Subject: [et-mgmt-tools] Bootstrapping cobbler from a DVD image In-Reply-To: <46042057.2080308@redhat.com> References: <46042057.2080308@redhat.com> Message-ID: <460424DF.20906@gmail.com> Michael DeHaan wrote: > I just wrote this up... > http://et.redhat.com/page/Cobbler_Import > > This basically shows how you can go from "blank system" with no > cobbler installed to network boot server really really quickly without > having to understand a lot of > the details of cobbler. > > Basically you can set up a boot server, from scratch, in about 20 > minutes or so -- and that includes waiting for the files to copy around. > > More articles/howtos to come over time, I'm sure. > > --Michael > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > Hi, Nice tutorial, but I think there is one command incomplete. That for a xen guest: * koan --server=bootserver.example.com --profile=FC-6-xen should be: * koan --virt --server=bootserver.example.com --profile=FC-6-xen Grtzz, Raf -------------- next part -------------- An HTML attachment was scrubbed... URL: From mdehaan at redhat.com Fri Mar 23 19:27:31 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 23 Mar 2007 15:27:31 -0400 Subject: [et-mgmt-tools] Bootstrapping cobbler from a DVD image In-Reply-To: <460424DF.20906@gmail.com> References: <46042057.2080308@redhat.com> <460424DF.20906@gmail.com> Message-ID: <46042A23.8080405@redhat.com> Raf Nijskens wrote: > Michael DeHaan wrote: >> I just wrote this up... >> http://et.redhat.com/page/Cobbler_Import >> >> This basically shows how you can go from "blank system" with no >> cobbler installed to network boot server really really quickly >> without having to understand a lot of >> the details of cobbler. >> >> Basically you can set up a boot server, from scratch, in about 20 >> minutes or so -- and that includes waiting for the files to copy around. >> >> More articles/howtos to come over time, I'm sure. >> >> --Michael >> >> _______________________________________________ >> et-mgmt-tools mailing list >> et-mgmt-tools at redhat.com >> https://www.redhat.com/mailman/listinfo/et-mgmt-tools >> > Hi, > > Nice tutorial, but I think there is one command incomplete. That for a > xen guest: > > * koan --server=bootserver.example.com --profile=FC-6-xen > > should be: > > * koan --virt --server=bootserver.example.com --profile=FC-6-xen > > > Grtzz, > > Raf > > > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools Updated. Thanks. From mdehaan at redhat.com Fri Mar 23 20:17:44 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 23 Mar 2007 16:17:44 -0400 Subject: [et-mgmt-tools] So soon? Cobbler 0.4.5 and koan 0.2.8 -- removal of unused "--virt-name" parameter from app. Message-ID: <460435E8.2030600@redhat.com> This release is rather inconsequental, but with the virtualization improvements included in koan 0.2.7, cobbler doesn't have any reason to have a "--virt-name" parameter anymore. In the future, koan may get a flag to override the default name koan creates, which is, right now, set to the mac address of the system. In order to remove this parameter from cobbler, one must update koan so that it no longer looks for it. Updating just cobbler /will/ break your systems with koan installed on them, so unfortunately both must be upgraded when moving to cobbler 0.4.5. If someone already was playing with cobbler 0.4.4, there is no rush to upgrade either, just keep in mind that the "--virt-name" parameter mentioned in the docs doesn't do anything :) --Michael From fj7716hz at aa.jp.fujitsu.com Mon Mar 26 00:50:03 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Mon, 26 Mar 2007 09:50:03 +0900 Subject: [et-mgmt-tools] [PATCH] HVM 3rd disk name don't use "hdc" In-Reply-To: <20070322152616.GD22912@redhat.com> References: <4602673D.8060802@aa.jp.fujitsu.com> <20070322152616.GD22912@redhat.com> Message-ID: <460718BB.7040602@aa.jp.fujitsu.com> Hi, Dan Daniel P. Berrange wrote: >> I try that making the HVM domain with 3+ disks by the virt-install command, >> but the command says following error message: >> # virt-install --name=HVM_3DISKS --file=/dev/sda6 --file=/dev/sda7 \ >> --file=/dev/sda8 --hvm --vnc --ram=376 --cdrom=/dev/cdrom > > This is a ticky one ! We need to a) refuse to allow > 3 --file arguments > to be used with HVM guests, and b) make sure the 3rd file gets given > hdd as its device name. Thank you for your reviewing. I rewite the patch. Signed-off-by: Tatsuro Enokura Thanks, Tatsuro Enokura -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: diskcheck.patch URL: From takatom at jp.fujitsu.com Mon Mar 26 11:09:26 2007 From: takatom at jp.fujitsu.com (Takahashi Tomohiro) Date: Mon, 26 Mar 2007 20:09:26 +0900 Subject: [et-mgmt-tools] [PATCH][RFC] Automatically pick the correct keymap based on the X keymap Message-ID: <01e301c76f97$3bb98fb0$ecb1220a@fj98229700> Hi, I made a patch that is automatically pick the correct keymap based on the X keymap. The virt-manager gets keymap information from "/etc/sysconfig/keyboard" and keytable.py (keytable.py is new python file.) For example. ============================= "/etc/sysconfig/keyboard" KEYBOARDTYPE="pc" KEYTABLE="jp106" ============================= If there is KEYTABLE="jp106", it compares "jp106"(/etc/sysconfig/keyboard) to "jp106"(keytable.py). and it set keymap = "ja". because "jp106"(keytable.py) is "ja" Other than those above. if KEYTABLE="us", it set keymap = "en-us". Because "us"(keytable.py) is "en-us" If there isn't "/etc/sysconfig/keyboard" or it cannot open "/etc/sysconfig/keyboard" The virt-manager doesn't get keymap information (same as before) and outputs error messeage at virt-manager.log keytable.py is incomplete file (ex."":"et"). Because I didn't know it. So, I think that if someone knows it, they support keytable.py Please give me an advice, if you have it. Signed-off-by: Tomohiro Takahashi Thanks, Tomohiro Takahashi. diff -uNrp virt-manager--devel.orig/src/virtManager/create.py virt-manager--devel/src/virtManager/create.py --- virt-manager/virt-manager--devel.orig/src/virtManager/create.py 2007-03-26 18:39:05.000000000 +0900 +++ virt-manager--devel/src/virtManager/create.py 2007-03-26 19:21:19.000000000 +0900 @@ -565,7 +565,24 @@ class vmmCreate(gobject.GObject): raise ValueError, "Unsupported networking type " + net[0] # set up the graphics to use SDL - guest.graphics = "vnc" + import keytable + keymap = None + vncport = None + try: + f = open("/etc/sysconfig/keyboard", "r") + except IOError, e: + logging.debug('Could not open "/etc/sysconfig/keyboard" ' + str(e)) + else: + while 1: + s = f.readline() + if s == "": + break + if re.search("KEYTABLE", s) != None: + kt = s.split('"')[1] + if keytable.keytable.has_key(kt): + keymap = keytable.keytable[kt] + f.close + guest.graphics = (True, "vnc", vncport, keymap) logging.debug("Creating a VM " + guest.name + \ "\n Type: " + guest.type + \ diff -uNrp virt-manager--devel.orig/src/virtManager/keytable.py virt-manager--devel/src/virtManager/keytable.py --- virt-manager--devel.orig/src/virtManager/keytable.py 1970-01-01 09:00:00.000000000 +0900 +++ virt-manager--devel/src/virtManager/keytable.py 2007-03-26 16:49:06.000000000 +0900 @@ -0,0 +1,37 @@ +# /etc/sysconfig/keyboard:keymap_name_for_qemu-dm +keytable = { + "ar":"ar", + "dk":"da", + "de":"de", + "de-ch":"de-ch", + "gb":"en-gb", + "us":"en-us", + "es":"es", + "":"et", + "se_FI":"fi", + "":"fo", + "":"fr", + "be":"fr-be", + "":"fr-ca", + "fr_CH":"fr_ch", + "hr":"hr", + "":"hu", + "":"is", + "it":"it", + "jp106":"ja", + "lt":"lt", + "lv":"lv", + "mk":"mk", + "": "nl", + "":"nl-be", + "no":"no", + "pl":"pl", + "pt":"pt", + "br":"pt-br", + "ru":"ru", + "sl":"sl", + "":"sv", + "th":"th", + "tr":"tr", + } + From fj7716hz at aa.jp.fujitsu.com Mon Mar 26 11:34:45 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Mon, 26 Mar 2007 20:34:45 +0900 Subject: [et-mgmt-tools] [PATCH] [RESEND]Check the making domain's mac address(part 2) In-Reply-To: <45FE5301.3020300@aa.jp.fujitsu.com> References: <45FE5301.3020300@aa.jp.fujitsu.com> Message-ID: <4607AFD5.7030107@aa.jp.fujitsu.com> Hi Would you give me a comment on this patch? If not, please apply it. Tatsuro Enokura wrote: > I made the patch the running domain's mac address check(changeset 125), > I make the patch in this time that the virt-install comand adds the > confirming message. > > The attached patch the following way: > > 1) Count that the making domian's MAC address is already use > the exists domain's MAC address. > 2) Check the hots's NIC MAC address. > 2) if count > 0 or , print confirmation message. > 3) if type yes then leave disk source as it is, > else ask disk source again. Signed-off-by: Tatsuro Enokura Thanks, Tatsuro Enokura -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: maccheck.patch URL: From hbrock at redhat.com Mon Mar 26 16:11:15 2007 From: hbrock at redhat.com (Hugh Brock) Date: Mon, 26 Mar 2007 12:11:15 -0400 Subject: [et-mgmt-tools] [PATCH] [RESEND]Check the making domain's mac address(part 2) In-Reply-To: <4607AFD5.7030107@aa.jp.fujitsu.com> References: <45FE5301.3020300@aa.jp.fujitsu.com> <4607AFD5.7030107@aa.jp.fujitsu.com> Message-ID: <4607F0A3.8040307@redhat.com> Tatsuro Enokura wrote: > Hi > > Would you give me a comment on this patch? > If not, please apply it. > > Tatsuro Enokura wrote: >> I made the patch the running domain's mac address check(changeset 125), >> I make the patch in this time that the virt-install comand adds the >> confirming message. >> >> The attached patch the following way: >> >> 1) Count that the making domian's MAC address is already use >> the exists domain's MAC address. >> 2) Check the hots's NIC MAC address. >> 2) if count > 0 or , print confirmation message. >> 3) if type yes then leave disk source as it is, >> else ask disk source again. > > Signed-off-by: Tatsuro Enokura > > Thanks, > Tatsuro Enokura > Hi. I had a look at the patch. The idea is OK, but there are two problems: first, I'd rather you not use exceptions for flow control; when you're testing for the mac address, if there's a problem, go straight to the user check rather than going through the whole try: except: sequence. Second problem is when I try it with a pre-existing mac address from the command line for an inactive, I no longer get a warning of a duplicate (which you set up in your last patch). If you can check these things and resubmit, I'd be happy to take it. Thanks, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From mdehaan at redhat.com Mon Mar 26 21:18:49 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 26 Mar 2007 17:18:49 -0400 Subject: [et-mgmt-tools] Cobbler and Koan moving to git Message-ID: <460838B9.9010603@redhat.com> After having some weird mercurial merge issues while working on virt-factory, I've decided to move cobbler and koan to git also. git is a version control system a lot like Mercurial, and has a mercurial like frontend called "cogito" (cg). You can see the shiny web interface at: http://git.et.redhat.com You can install git as follows: yum install git cogito To get a checkout of the cobbler and koan sources: cg clone git://et.redhat.com/git/cobbler cg clone git://et.redhat.com/git/koan To get updates as they occur: cd $projectdir cg update If anyone has any problems with the repository, let us know. I'll update the http://cobbler.et.redhat.com/ webpage shortly to point to the new repository. The mercurial repos for cobbler & koan will probably be decommissioned soon also, to reduce confusion. Thanks to Jim Meyering for a lot of help in setting this server up and getting stuff moved over. --Michael DeHaan From takatom at jp.fujitsu.com Tue Mar 27 04:12:08 2007 From: takatom at jp.fujitsu.com (Takahashi Tomohiro) Date: Tue, 27 Mar 2007 13:12:08 +0900 Subject: [et-mgmt-tools] [PATCH] Automatically pick the correct keymap based on the X keymap Message-ID: <001201c77026$1d56e290$ecb1220a@fj98229700> Hi, I made a patch that is automatically pick the correct keymap based on the X keymap. The virt-manager gets keymap information from "/etc/sysconfig/keyboard" and keytable.py (keytable.py is new python file.) For example. ============================= "/etc/sysconfig/keyboard" KEYBOARDTYPE="pc" KEYTABLE="jp106" ============================= If there is KEYTABLE="jp106", it compares "jp106"(/etc/sysconfig/keyboard) to "jp106"(keytable.py). and it set keymap = "ja". because "jp106"(keytable.py) is "ja" Other than those above. if KEYTABLE="us", it set keymap = "en-us". Because "us"(keytable.py) is "en-us" If there isn't "/etc/sysconfig/keyboard" or it cannot open "/etc/sysconfig/keyboard" The virt-manager doesn't get keymap information (same as before) and outputs error messeage at virt-manager.log keytable.py is incomplete file (ex."":"et"). Because I didn't know it. So, I think that if someone knows it, they support keytable.py Please give me an advice, if you have it. Signed-off-by: Tomohiro Takahashi Thanks, Tomohiro Takahashi. diff -uNrp virt-manager--devel.orig/src/virtManager/create.py virt-manager--devel/src/virtManager/create.py --- virt-manager--devel.orig/src/virtManager/create.py 2007-03-26 18:39:05.000000000 +0900 +++ virt-manager--devel/src/virtManager/create.py 2007-03-27 11:04:05.000000000 +0900 @@ -55,6 +55,8 @@ PAGE_NETWORK = 6 PAGE_CPUMEM = 7 PAGE_SUMMARY = 8 +KEYBOARD_DIR = "/etc/sysconfig/keyboard" + class vmmCreateMeter(progress.BaseMeter): def __init__(self, asyncjob): # progress meter has to run asynchronously, so pass in the @@ -565,7 +567,24 @@ class vmmCreate(gobject.GObject): raise ValueError, "Unsupported networking type " + net[0] # set up the graphics to use SDL - guest.graphics = "vnc" + import keytable + keymap = None + vncport = None + try: + f = open(KEYBOARD_DIR, "r") + except IOError, e: + logging.debug('Could not open "/etc/sysconfig/keyboard" ' + str(e)) + else: + while 1: + s = f.readline() + if s == "": + break + if re.search("KEYTABLE", s) != None: + kt = s.split('"')[1] + if keytable.keytable.has_key(kt): + keymap = keytable.keytable[kt] + f.close + guest.graphics = (True, "vnc", vncport, keymap) logging.debug("Creating a VM " + guest.name + \ "\n Type: " + guest.type + \ diff -uNrp virt-manager--devel.orig/src/virtManager/keytable.py virt-manager--devel/src/virtManager/keytable.py --- virt-manager--devel.orig/src/virtManager/keytable.py 1970-01-01 09:00:00.000000000 +0900 +++ virt-manager--devel/src/virtManager/keytable.py 2007-03-26 16:49:06.000000000 +0900 @@ -0,0 +1,37 @@ +# keytable_name_from_dom0:keymap_name_to_Xen +keytable = { + "ar":"ar", + "dk":"da", + "de":"de", + "de-ch":"de-ch", + "gb":"en-gb", + "us":"en-us", + "es":"es", + "":"et", + "se_FI":"fi", + "":"fo", + "":"fr", + "be":"fr-be", + "":"fr-ca", + "fr_CH":"fr_ch", + "hr":"hr", + "":"hu", + "":"is", + "it":"it", + "jp106":"ja", + "lt":"lt", + "lv":"lv", + "mk":"mk", + "": "nl", + "":"nl-be", + "no":"no", + "pl":"pl", + "pt":"pt", + "br":"pt-br", + "ru":"ru", + "sl":"sl", + "":"sv", + "th":"th", + "tr":"tr", + } + From fj7716hz at aa.jp.fujitsu.com Tue Mar 27 04:26:10 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Tue, 27 Mar 2007 13:26:10 +0900 Subject: [et-mgmt-tools] [PATCH] [RESEND]Check the making domain's mac address(part 2) In-Reply-To: <4607F0A3.8040307@redhat.com> References: <45FE5301.3020300@aa.jp.fujitsu.com> <4607AFD5.7030107@aa.jp.fujitsu.com> <4607F0A3.8040307@redhat.com> Message-ID: <46089CE2.5000802@aa.jp.fujitsu.com> Hi Hugh Hugh Brock wrote: > Hi. I had a look at the patch. The idea is OK, but there are two > problems: first, I'd rather you not use exceptions for flow control; > when you're testing for the mac address, if there's a problem, go > straight to the user check rather than going through the whole try: > except: sequence. Thank you for youre reviewing. I rewrite a patch. > Second problem is when I try it with a pre-existing > mac address from the command line for an inactive, I no longer get a > warning of a duplicate (which you set up in your last patch). I assume User run various domains with specified time. For this case, I just show warning messages only. Since user intent to run the same mac address with various domain in various time. Thanks Tatsuro Enokura -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: maccheck.patch URL: From y.yaegashi at jp.fujitsu.com Tue Mar 27 04:34:32 2007 From: y.yaegashi at jp.fujitsu.com (Yoshihiko Yaegashi) Date: Tue, 27 Mar 2007 13:34:32 +0900 Subject: [et-mgmt-tools] [RFC] virt-manager test by the dogtail Message-ID: <46089ED8.2030901@jp.fujitsu.com> Hi, I sent a mail to "dogtail-devel-list", but I hardly ever get back a reply. So I send a mail to this mailing list. I test virt-manager by the dogtail. I cannot make a 'table cell' active. In the following descriptions of the attached file, because activate('Domain-0', roleName='table cell') cannot make a 'table cell' active, "Open" button is not sensitive. ... activate('Domain-0', roleName='table cell') click('Open', roleName='push button') ... Can I ask your advice on something? N.B. Please convert "localhost" into "your server name" within the sample code of the attached file. Thanks Yaegashi -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: test.py URL: From fj0873gn at aa.jp.fujitsu.com Tue Mar 27 06:20:33 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Tue, 27 Mar 2007 15:20:33 +0900 Subject: [et-mgmt-tools] [PATCH][RESEND] RuntimeError message is error In-Reply-To: <200703191901.CAB48982.40O8GHK9@aa.jp.fujitsu.com> References: <200703191901.CAB48982.40O8GHK9@aa.jp.fujitsu.com> Message-ID: <200703271520.GAH78136.K8O4HG09@aa.jp.fujitsu.com> Hi Would you give me a comment on this patch? If not, please apply it. > The RuntimeError message becomes the error > "NameError: global name 'os_type' is not defined". > > The attached patch adds to solve this problem. > > Signed-off-by: Nobuhiro Itou Thanks, Nobuhiro Itou. --- a/virtinst/FullVirtGuest.py Fri Mar 23 10:10:03 2007 -0400 +++ b/virtinst/FullVirtGuest.py Tue Mar 27 15:14:52 2007 +0900 @@ -72,7 +72,7 @@ class FullVirtGuest(Guest.XenGuest): if FullVirtGuest.OS_TYPES[self._os_type].has_key(val): self._os_variant = val else: - raise RuntimeError, "OS variant %s does not exist in our dictionary for OS type %s" % (val, os_type) + raise RuntimeError, "OS variant %s does not exist in our dictionary for OS type %s" % (val, self._os_type) os_variant = property(get_os_variant, set_os_variant) def set_os_type_parameters(self, os_type, os_variant): From hbrock at redhat.com Tue Mar 27 12:10:28 2007 From: hbrock at redhat.com (Hugh Brock) Date: Tue, 27 Mar 2007 08:10:28 -0400 Subject: [et-mgmt-tools] [RFC] virt-manager test by the dogtail In-Reply-To: <46089ED8.2030901@jp.fujitsu.com> References: <46089ED8.2030901@jp.fujitsu.com> Message-ID: <460909B4.2050700@redhat.com> Yoshihiko Yaegashi wrote: > Hi, > > I sent a mail to "dogtail-devel-list", but I hardly ever > get back a reply. > So I send a mail to this mailing list. > > I test virt-manager by the dogtail. > I cannot make a 'table cell' active. > > In the following descriptions of the attached file, > because activate('Domain-0', roleName='table cell') cannot make > a 'table cell' active, "Open" button is not sensitive. > > ... > activate('Domain-0', roleName='table cell') > click('Open', roleName='push button') > ... > > Can I ask your advice on something? > > N.B. Please convert "localhost" into "your server name" within > the sample code of the attached file. > > Thanks > Yaegashi > Very interesting, I will track down a dogtail expert and try to figure out what is going on here. Thanks! --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From hbrock at redhat.com Tue Mar 27 12:14:59 2007 From: hbrock at redhat.com (Hugh Brock) Date: Tue, 27 Mar 2007 08:14:59 -0400 Subject: [et-mgmt-tools] [PATCH][RESEND] RuntimeError message is error In-Reply-To: <200703271520.GAH78136.K8O4HG09@aa.jp.fujitsu.com> References: <200703191901.CAB48982.40O8GHK9@aa.jp.fujitsu.com> <200703271520.GAH78136.K8O4HG09@aa.jp.fujitsu.com> Message-ID: <46090AC3.3040708@redhat.com> Nobuhiro Itou wrote: > Hi > > Would you give me a comment on this patch? > If not, please apply it. > >> The RuntimeError message becomes the error >> "NameError: global name 'os_type' is not defined". >> >> The attached patch adds to solve this problem. >> >> Signed-off-by: Nobuhiro Itou > > > Thanks, > Nobuhiro Itou. > > So sorry, I thought I had applied this already! Applied and pushed as of this moment. Thanks! --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From hbrock at redhat.com Tue Mar 27 15:43:49 2007 From: hbrock at redhat.com (Hugh Brock) Date: Tue, 27 Mar 2007 11:43:49 -0400 Subject: [et-mgmt-tools] [PATCH] [RESEND]Check the making domain's mac address(part 2) In-Reply-To: <46089CE2.5000802@aa.jp.fujitsu.com> References: <45FE5301.3020300@aa.jp.fujitsu.com> <4607AFD5.7030107@aa.jp.fujitsu.com> <4607F0A3.8040307@redhat.com> <46089CE2.5000802@aa.jp.fujitsu.com> Message-ID: <46093BB5.5080000@redhat.com> Tatsuro Enokura wrote: > Hi Hugh > > Hugh Brock wrote: >> Hi. I had a look at the patch. The idea is OK, but there are two >> problems: first, I'd rather you not use exceptions for flow control; >> when you're testing for the mac address, if there's a problem, go >> straight to the user check rather than going through the whole try: >> except: sequence. > > Thank you for youre reviewing. > I rewrite a patch. > >> Second problem is when I try it with a pre-existing >> mac address from the command line for an inactive, I no longer get a >> warning of a duplicate (which you set up in your last patch). > > I assume User run various domains with specified time. > For this case, I just show warning messages only. > Since user intent to run the same mac address > with various domain in various time. > > Thanks > Tatsuro Enokura > Actually now that I have thought some more about this, I don't think we want to apply it at all. The current semantics of the --mac flag are as follows: 1. If --mac is specified, attempt to use that address. If it conflicts with an in-use address, fail with an error 2. If --mac is not specified, choose an unused mac address at random. You would like to add the case wherein if --mac is specified and conflicts with an existing mac address, a user can override the conflict. However, we want operations with virt-install to be scriptable -- that is, we do not ever want the script to halt waiting for input unless it is obviously being run interactively. This means that, since a mac address is never entered at a prompt, we can't put up a warning if it conflicts. And I don't think we want to change the semantics to always prompt for a mac address if it is not specified, since most users aren't going to care what the mac address is anyway. One way we could address the underlying problem is to change the mac-address conflict checking code in VirtualNetworkInterface.setup() so that it prints a warning to the console or logs it, but continues with the guest creation. I'm fine with this behavior, since it seems reasonable to me that an installer might want to have multiple inactive guests with the same mac address. Another alternative would be to fail if there is an active guest with the same mac address, but only print a warning (and continue) if there is an inactive guest with the same mac address. I would take a patch for either scenario. Thanks, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From dhawal at netmagicsolutions.com Tue Mar 27 16:07:22 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Tue, 27 Mar 2007 21:37:22 +0530 Subject: [et-mgmt-tools] cobbler_syslogd question Message-ID: <4609413A.7010800@netmagicsolutions.com> Hello List / Michael, # ps -ef | grep cob root 2668 1 0 21:20 ? 00:00:00 /usr/bin/python /usr/bin/cobbler_syslogd # netstat -nlp | grep 2668 udp 0 0 202.87.39.87:25150 0.0.0.0:* 2668/python Why is cobbler_syslogd listening only on loopback? how do i make it listen to all inets? - dhawal From dhawal at netmagicsolutions.com Tue Mar 27 16:11:18 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Tue, 27 Mar 2007 21:41:18 +0530 Subject: [et-mgmt-tools] cobbler_syslogd question In-Reply-To: <4609413A.7010800@netmagicsolutions.com> References: <4609413A.7010800@netmagicsolutions.com> Message-ID: <46094226.6020400@netmagicsolutions.com> Dhawal Doshy wrote: > Hello List / Michael, > > # ps -ef | grep cob > root 2668 1 0 21:20 ? 00:00:00 /usr/bin/python > /usr/bin/cobbler_syslogd > > # netstat -nlp | grep 2668 > udp 0 0 202.87.39.87:25150 0.0.0.0:* > 2668/python > > Why is cobbler_syslogd listening only on loopback? how do i make it > listen to all inets? Can i recall this question?? please.. :-( correcting /etc/hosts resolved the error. From dhawal at netmagicsolutions.com Tue Mar 27 16:47:43 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Tue, 27 Mar 2007 22:17:43 +0530 Subject: [et-mgmt-tools] cobbler_syslogd question In-Reply-To: <46094226.6020400@netmagicsolutions.com> References: <4609413A.7010800@netmagicsolutions.com> <46094226.6020400@netmagicsolutions.com> Message-ID: <46094AAF.4050301@netmagicsolutions.com> Dhawal Doshy wrote: > Dhawal Doshy wrote: >> Hello List / Michael, >> >> # ps -ef | grep cob >> root 2668 1 0 21:20 ? 00:00:00 /usr/bin/python >> /usr/bin/cobbler_syslogd >> >> # netstat -nlp | grep 2668 >> udp 0 0 202.87.39.87:25150 0.0.0.0:* >> 2668/python >> >> Why is cobbler_syslogd listening only on loopback? how do i make it >> listen to all inets? > > Can i recall this question?? please.. :-( > correcting /etc/hosts resolved the error. Umm more junk from me.. question remains.. how do i make it listen on 0.0.0.0 rather than a specific interface.. newer servers are connected over a cross-cable during installation (eth1) which remains link-down until provisioning is started. From mdehaan at redhat.com Tue Mar 27 17:19:12 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 27 Mar 2007 13:19:12 -0400 Subject: [et-mgmt-tools] cobbler_syslogd question In-Reply-To: <46094AAF.4050301@netmagicsolutions.com> References: <4609413A.7010800@netmagicsolutions.com> <46094226.6020400@netmagicsolutions.com> <46094AAF.4050301@netmagicsolutions.com> Message-ID: <46095210.1040704@redhat.com> Dhawal Doshy wrote: > Dhawal Doshy wrote: >> Dhawal Doshy wrote: >>> Hello List / Michael, >>> >>> # ps -ef | grep cob >>> root 2668 1 0 21:20 ? 00:00:00 /usr/bin/python >>> /usr/bin/cobbler_syslogd >>> >>> # netstat -nlp | grep 2668 >>> udp 0 0 202.87.39.87:25150 0.0.0.0:* >>> 2668/python >>> >>> Why is cobbler_syslogd listening only on loopback? how do i make it >>> listen to all inets? >> >> Can i recall this question?? please.. :-( >> correcting /etc/hosts resolved the error. > > Umm more junk from me.. question remains.. how do i make it listen on > 0.0.0.0 rather than a specific interface.. newer servers are connected > over a cross-cable during installation (eth1) which remains link-down > until provisioning is started. > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools It looks like syslogd was being a bit stupid about what it tried to bind to... See if this helps you out, and if so, I'll include it in the next release. diff --git a/cobbler/syslog_watcher.py b/cobbler/syslog_watcher.py index 64dd802..4846b3e 100644 --- a/cobbler/syslog_watcher.py +++ b/cobbler/syslog_watcher.py @@ -22,10 +22,11 @@ def main(): port = settings.syslog_port s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM) - try: - s.bind((socket.gethostname(), port)) - except: - s.bind("127.0.0.1", port)) + #try: + # s.bind((socket.gethostname(), port)) + #except: + # s.bind("127.0.0.1", port)) + s.bind(("0.0.0.0", port)) buf = 1024 --Michael From dhawal at netmagicsolutions.com Tue Mar 27 17:29:00 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Tue, 27 Mar 2007 22:59:00 +0530 Subject: [et-mgmt-tools] cobbler_syslogd question In-Reply-To: <46095210.1040704@redhat.com> References: <4609413A.7010800@netmagicsolutions.com> <46094226.6020400@netmagicsolutions.com> <46094AAF.4050301@netmagicsolutions.com> <46095210.1040704@redhat.com> Message-ID: <4609545C.4080805@netmagicsolutions.com> Michael DeHaan wrote: > Dhawal Doshy wrote: >> Dhawal Doshy wrote: >>> Dhawal Doshy wrote: >>>> Hello List / Michael, >>>> >>>> # ps -ef | grep cob >>>> root 2668 1 0 21:20 ? 00:00:00 /usr/bin/python >>>> /usr/bin/cobbler_syslogd >>>> >>>> # netstat -nlp | grep 2668 >>>> udp 0 0 202.87.39.87:25150 0.0.0.0:* >>>> 2668/python >>>> >>>> Why is cobbler_syslogd listening only on loopback? how do i make it >>>> listen to all inets? >>> >>> Can i recall this question?? please.. :-( >>> correcting /etc/hosts resolved the error. >> >> Umm more junk from me.. question remains.. how do i make it listen on >> 0.0.0.0 rather than a specific interface.. newer servers are connected >> over a cross-cable during installation (eth1) which remains link-down >> until provisioning is started. > > It looks like syslogd was being a bit stupid about what it tried to bind > to... > > See if this helps you out, and if so, I'll include it in the next release. > > > diff --git a/cobbler/syslog_watcher.py b/cobbler/syslog_watcher.py > index 64dd802..4846b3e 100644 > --- a/cobbler/syslog_watcher.py > +++ b/cobbler/syslog_watcher.py > @@ -22,10 +22,11 @@ def main(): > port = settings.syslog_port > > s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM) > - try: > - s.bind((socket.gethostname(), port)) > - except: > - s.bind("127.0.0.1", port)) > + #try: > + # s.bind((socket.gethostname(), port)) > + #except: > + # s.bind("127.0.0.1", port)) > + s.bind(("0.0.0.0", port)) > > buf = 1024 Cool thanks, works as intended.. (i really really need to learn python) From trevor.carpenter at ed.ac.uk Tue Mar 27 17:41:32 2007 From: trevor.carpenter at ed.ac.uk (Trevor Carpenter) Date: Tue, 27 Mar 2007 18:41:32 +0100 Subject: [et-mgmt-tools] virtualisation tooling - number crunchers Message-ID: <1175017292.315.22.camel@localhost.localdomain> Hi, We have a dell Power edge server running RH 4 ES. It has 4 processors each with dual cores, 8Gb of ram and 4 146Gb disks (currently raided). We want to be able to set this up to 'appear' as 4 separate machines each running a separate instance of RH with a resource of 146Gb disk (or a quarter of whats left), 2Gb of ram (or a quarter of whats left) and 1 processor each. Ideally I would like them to have there own IP address. We want to use them to run console number crunching applications whilst ensuring that each 'machine' does not take more than it's fair share of resource. We do not need the X server to be running and will connect via ssh if appropriate. Is the virtualisation tooling described here the way to do this? Or is there another mechanism available already built into the OS (I worked in an environment with Solaris 10, this seems as close as I can get to the containers we used there.). I will happily convert the raid array drives to Jbod. I am reading the documentation, and I am not asking for 'sys admin' or 'support', I just want to confirm the Virtual Machine Manager etc is the correct approach. I can confirm more about the chipset/hardware if required. To understand more about what we do please visit: http://www.sbirc.ed.ac.uk/research/research.asp we are not a commercial organization. Thank you in advance for any help you can provide. Trevor. From jim at rossberry.com Tue Mar 27 19:59:38 2007 From: jim at rossberry.com (Jim Wildman) Date: Tue, 27 Mar 2007 15:59:38 -0400 (EDT) Subject: [et-mgmt-tools] cobbler_syslogd question In-Reply-To: <46095210.1040704@redhat.com> References: <4609413A.7010800@netmagicsolutions.com> <46094226.6020400@netmagicsolutions.com> <46094AAF.4050301@netmagicsolutions.com> <46095210.1040704@redhat.com> Message-ID: Well, I was going to submit this as a separate bug, but since you posted it here.. On Tue, 27 Mar 2007, Michael DeHaan wrote: > diff --git a/cobbler/syslog_watcher.py b/cobbler/syslog_watcher.py > index 64dd802..4846b3e 100644 > --- a/cobbler/syslog_watcher.py > +++ b/cobbler/syslog_watcher.py > @@ -22,10 +22,11 @@ def main(): > port = settings.syslog_port > > s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM) > - try: > - s.bind((socket.gethostname(), port)) > - except: > - s.bind("127.0.0.1", port)) > + #try: > + # s.bind((socket.gethostname(), port)) > + #except: > + # s.bind("127.0.0.1", port)) ^^ one of these must die > + s.bind(("0.0.0.0", port)) > > buf = 1024 > > > --Michael > > > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > ------------------------------------------------------------------------ Jim Wildman, CISSP, RHCE jim at rossberry.com http://www.rossberry.com "Society in every state is a blessing, but Government, even in its best state, is a necessary evil; in its worst state, an intolerable one." Thomas Paine From mdehaan at redhat.com Tue Mar 27 20:47:44 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 27 Mar 2007 16:47:44 -0400 Subject: [et-mgmt-tools] cobbler_syslogd question In-Reply-To: References: <4609413A.7010800@netmagicsolutions.com> <46094226.6020400@netmagicsolutions.com> <46094AAF.4050301@netmagicsolutions.com> <46095210.1040704@redhat.com> Message-ID: <460982F0.8050606@redhat.com> Jim Wildman wrote: > Well, I was going to submit this as a separate bug, but since you posted > it here.. > > On Tue, 27 Mar 2007, Michael DeHaan wrote: > >> diff --git a/cobbler/syslog_watcher.py b/cobbler/syslog_watcher.py >> index 64dd802..4846b3e 100644 >> --- a/cobbler/syslog_watcher.py >> +++ b/cobbler/syslog_watcher.py >> @@ -22,10 +22,11 @@ def main(): >> port = settings.syslog_port >> >> s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM) >> - try: >> - s.bind((socket.gethostname(), port)) >> - except: >> - s.bind("127.0.0.1", port)) >> + #try: >> + # s.bind((socket.gethostname(), port)) >> + #except: >> + # s.bind("127.0.0.1", port)) > ^^ one of these must die >> + s.bind(("0.0.0.0", port)) >> >> buf = 1024 >> >> >> --Michael >> >> >> >> >> _______________________________________________ >> et-mgmt-tools mailing list >> et-mgmt-tools at redhat.com >> https://www.redhat.com/mailman/listinfo/et-mgmt-tools >> > > ------------------------------------------------------------------------ > Jim Wildman, CISSP, RHCE jim at rossberry.com http://www.rossberry.com > "Society in every state is a blessing, but Government, even in its best > state, is a necessary evil; in its worst state, an intolerable one." > Thomas Paine > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools Bad testing on my part -- those running 0.4.5 andwanting syslog to work, note the missing paren. Preferably just apply the "0.0.0.0" fix... --Michael From matts at hostingsupport.com Tue Mar 27 21:59:52 2007 From: matts at hostingsupport.com (Matt S [UNIX Administrator]) Date: Tue, 27 Mar 2007 16:59:52 -0500 Subject: [et-mgmt-tools] Distro specific kickstarts and repo questions In-Reply-To: <46042057.2080308@redhat.com> References: <46042057.2080308@redhat.com> Message-ID: <1175032793.5768.20.camel@shire.propagation.net> Would it be possible to track kickstarts based on distro? Alternatively could we set a default kickstart on a per-distro basis. In our environment we offer a monolithic configuration on various sets of hardware so it would be advantageous if we could use the built-in "menu" functionality and expect that the default kickstart will be appropriate for our operating system selection. Maybe /etc/cobbler/$distro/default.ks For instance, we may not want the default kickstart to do things in post on Fedora that we want in RHEL-5. Perhaps there is a way to manage this in cobbler and I'm just missing it. Any thoughts? (I'm trying to accomplish this somewhat through the use of profiles) A second question. We have some custom repo's that we want to configure in order that packages may be installed during provisioning. I setup a mirrored repo which imported all the files correctly. When the repo is defined during the profile creation it works but cobbler won't sync. /var/lib/cobbler/profiles shows the repo listed The following error is kicked back: one of the listed repositories is not defined in cobbler Any help would be appreciated. From mdehaan at redhat.com Tue Mar 27 22:27:58 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 27 Mar 2007 18:27:58 -0400 Subject: [et-mgmt-tools] Distro specific kickstarts and repo questions In-Reply-To: <1175032793.5768.20.camel@shire.propagation.net> References: <46042057.2080308@redhat.com> <1175032793.5768.20.camel@shire.propagation.net> Message-ID: <46099A6E.8090108@redhat.com> Matt S [UNIX Administrator] wrote: > Would it be possible to track kickstarts based on distro? Alternatively > could we set a default kickstart on a per-distro basis. In our > environment we offer a monolithic configuration on various sets of > hardware so it would be advantageous if we could use the built-in "menu" > functionality and expect that the default kickstart will be appropriate > for our operating system selection. What you want to do is create a profile for the distribution... The distros, by definition, do not contain kickstart information -- this is by design. While you may have a monolithic setup now, once you start wanting to provision two or three different types of machines that all run FC-6-i386 (for instance), the benefits of having different profiles becomes readily apparent. Until then, it's just a minor inconvenience for having a good abstraction layer :) For example, you might have profiles for RHEL-5-dbserver, another for RHEL-5-appserver, and another profile for RHEL-5-webserver. These are both RHEL-5, but they are installed differently. We want folks to start thinking about the roles of specific machines rather than just what operating system they run, basically. The same concepts are, in many ways, what we try to get across when using Puppet. However, if you just want to have one profile per distro for starters, that's totally ok. > > Maybe /etc/cobbler/$distro/default.ks > > For instance, we may not want the default kickstart to do things in post > on Fedora that we want in RHEL-5. You can assign any kickstarts you want to any profiles. So this isn't a problem... cobbler profile add --name=FC-6-i386-general --distro=FC-6-i386 --distro=FC-6-i386 --kickstart=/opt/foo/mykickstart_fc6.ks cobbler profile add --name=RHEL-5-i386-general --distro=RHEL-5-i386 --distro=RHEL-5-i386 --kickstart=/opt/foo/mykickstart_rhel5.ks > > Perhaps there is a way to manage this in cobbler and I'm just missing > it. > > Any thoughts? (I'm trying to accomplish this somewhat through the use of > profiles) Yes, profiles are exactly what you should use. A distro describes what is being installed (the install source), a profile describes how it is being installed. > > > > A second question. We have some custom repo's that we want to configure > in order that packages may be installed during provisioning. I setup a > mirrored repo which imported all the files correctly. > > > When the repo is defined during the profile creation it works but > cobbler won't sync. > > /var/lib/cobbler/profiles shows the repo listed > > The following error is kicked back: one of the listed repositories is > not defined in cobbler What version of cobbler are you using? Also what does your /var/lib/cobbler/profiles look like? > > Any help would be appreciated. Sure, send me the above if you don't mind... > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From berrange at redhat.com Wed Mar 28 00:01:20 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 28 Mar 2007 01:01:20 +0100 Subject: [et-mgmt-tools] virt-manager new host overview & resources window Message-ID: <20070328000120.GG3126@redhat.com> I've just pushed a change to virt-manager to adds a new 'Host details' window. Thing of it as the equivalent of 'Guest details', but giving info about the host. It currently shows: - Hostname, hypervisor type, physical RAM, logical CPU count, CPU arch - Graphs of overall host CPU & memory usage (aggregate of all guest usage) - List of virtual networks & their configuration To come later - Ability to create new virtual networks - Ability to edit / delete virtual networks - Storage pools. In absence of libvirt having any storage APIs in the very short term I'm just going to hardcode a single storage pool '/var/lib/xen/images' or '/var/lib/qemu/images' or $HOME/.libvirt/qemu/images' as appropriate for the HV being used. Instead of having people enter an explicit path to create a disk image under, they'll just enter a name & select a storage pool & we'll automatically create it in the correct place. This will avoid all these SELinux issues with people picking random non-labelled directories. When libvirt gets formal storage APIs, we'll be able to create / change storage pools at will. Attaching two screenshots of current state Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| -------------- next part -------------- A non-text attachment was scrubbed... Name: host-summary.png Type: image/png Size: 26908 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: host-networks.png Type: image/png Size: 34304 bytes Desc: not available URL: From sakaia at jp.fujitsu.com Wed Mar 28 00:37:30 2007 From: sakaia at jp.fujitsu.com (Atsushi SAKAI) Date: Wed, 28 Mar 2007 09:37:30 +0900 Subject: [et-mgmt-tools] virt-manager new host overview & resources window In-Reply-To: <20070328000120.GG3126@redhat.com> References: <20070328000120.GG3126@redhat.com> Message-ID: <200703280038.l2S0c0Fm012374@fjmscan503.ms.jp.fujitsu.com> Hi, Dan Is there any plan to show Scheduler information details? (weight, cap etc) I am plan to make scheduler API on libvirt. or this issue should be discussed in libvirt? Thanks Atsushi SAKAI "Daniel P. Berrange" wrote: > I've just pushed a change to virt-manager to adds a new 'Host details' > window. Thing of it as the equivalent of 'Guest details', but giving info > about the host. It currently shows: From berrange at redhat.com Wed Mar 28 00:40:54 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 28 Mar 2007 01:40:54 +0100 Subject: [et-mgmt-tools] virt-manager new host overview & resources window In-Reply-To: <200703280038.l2S0c0Fm012374@fjmscan503.ms.jp.fujitsu.com> References: <20070328000120.GG3126@redhat.com> <200703280038.l2S0c0Fm012374@fjmscan503.ms.jp.fujitsu.com> Message-ID: <20070328004053.GH3126@redhat.com> On Wed, Mar 28, 2007 at 09:37:30AM +0900, Atsushi SAKAI wrote: > Hi, Dan > > Is there any plan to show Scheduler information details? > (weight, cap etc) > I am plan to make scheduler API on libvirt. > or this issue should be discussed in libvirt? If there are libvirt APIs, then we could think about how we would present the information to the user. So, yes, best to discuss any possible APIs on the libvir-list. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From fj7716hz at aa.jp.fujitsu.com Wed Mar 28 08:44:14 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Wed, 28 Mar 2007 17:44:14 +0900 Subject: [et-mgmt-tools] [PATCH] [RESEND]Check the making domain's mac address(part 2) In-Reply-To: <46093BB5.5080000@redhat.com> References: <45FE5301.3020300@aa.jp.fujitsu.com> <4607AFD5.7030107@aa.jp.fujitsu.com> <4607F0A3.8040307@redhat.com> <46089CE2.5000802@aa.jp.fujitsu.com> <46093BB5.5080000@redhat.com> Message-ID: <460A2ADE.5030007@aa.jp.fujitsu.com> Hi Hugh Hugh Brock wrote: > Actually now that I have thought some more about this, I don't think we > want to apply it at all. > > The current semantics of the --mac flag are as follows: > > 1. If --mac is specified, attempt to use that address. If it conflicts > with an in-use address, fail with an error > > 2. If --mac is not specified, choose an unused mac address at random. > > You would like to add the case wherein if --mac is specified and > conflicts with an existing mac address, a user can override the > conflict. However, we want operations with virt-install to be scriptable > -- that is, we do not ever want the script to halt waiting for input > unless it is obviously being run interactively. This means that, since a > mac address is never entered at a prompt, we can't put up a warning if > it conflicts. And I don't think we want to change the semantics to > always prompt for a mac address if it is not specified, since most users > aren't going to care what the mac address is anyway. > > One way we could address the underlying problem is to change the > mac-address conflict checking code in VirtualNetworkInterface.setup() so > that it prints a warning to the console or logs it, but continues with > the guest creation. I'm fine with this behavior, since it seems > reasonable to me that an installer might want to have multiple inactive > guests with the same mac address. Another alternative would be to fail > if there is an active guest with the same mac address, but only print a > warning (and continue) if there is an inactive guest with the same mac > address. I would take a patch for either scenario. I agree with your suggestion. I rewrite a patch that the mac-address conflict checking code in VirtualNetworkInterface.setup() so that it prints a warning to the stderr and logs. Thanks, Tatsuro Enokura -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: maccheck.patch URL: From fj0873gn at aa.jp.fujitsu.com Wed Mar 28 08:50:15 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Wed, 28 Mar 2007 17:50:15 +0900 Subject: [et-mgmt-tools] Subject:[PATCH][RESEND] Windows HVM can't continue toinstall In-Reply-To: <4602DC54.7050203@redhat.com> References: <200703201857.AEF73948.08KGO94H@aa.jp.fujitsu.com> <4602DC54.7050203@redhat.com> Message-ID: <200703281750.IGC73478.O9G80HK4@aa.jp.fujitsu.com> Hi Hugh > We don't really want to put OS-specific stuff into the virt-install code > if we can help it; we made an exception for the config flag setting > stuff because we're just abstracting behavior (setting acpi= and apic=) > that is already available. All right. In fact, I think OS-specific stuff shouldn't put into the virt-install, too. > However I don't see any reason why a patch that would allow a virtinst > argument like "--keep-cdrom" and then put something like " type="block" device="cdrom">... etc" in the domain's permanent xml > wouldn't work. This would mean that the domain would have access to the > cdrom drive every time it reboots. Of course, it also means we'll need > to provide some way to disconnect the domain from that drive once the > install process is finished, but we'll get to that later. > > Let me know if that works for you... I remaked a patch. This time, it has the flag in FullVirtGuest.OS_TYPES. The OS-specific code is removed from virt-inst. How about this correction? Signed-off-by: Nobuhiro Itou Thanks, Nobuhiro Itou. ------------------------------------------------------------- diff -r 36a9973c2e28 virt-install --- a/virt-install Tue Mar 27 08:13:38 2007 -0400 +++ b/virt-install Tue Mar 27 19:24:31 2007 +0900 @@ -535,6 +535,7 @@ def main(): if not hvm: # paravirt get_paravirt_install(options.location, guest) get_paravirt_extraargs(options.extra, guest) + continue_inst = False else: get_fullvirt_cdrom(options.cdrom, guest) guest.set_os_type(options.os_type) @@ -543,6 +544,7 @@ def main(): guest.features["acpi"] = False if options.noapic: guest.features["apic"] = False + continue_inst = guest.get_continue_inst() if options.autoconsole is False: conscb = None @@ -577,6 +579,17 @@ def main(): struct.unpack("H", buf[0x1fe: 0x200]) == (0xaa55,): # things installed enough that we should be able to restart # the domain + if continue_inst: + # continue to installation. + dom = guest.continue_install(conscb,progresscb) + if dom is None: + print "Guest installation failed" + sys.exit(0) + elif dom.info()[0] != libvirt.VIR_DOMAIN_SHUTOFF: + # domain seems to be running + print "Domain installation still in progress. You can reconnect " + print "to the console to complete the installation process." + sys.exit(0) print "Guest installation complete... restarting guest." dom.create() guest.connect_console(conscb) diff -r 36a9973c2e28 virtinst/Guest.py --- a/virtinst/Guest.py Tue Mar 27 08:13:38 2007 -0400 +++ b/virtinst/Guest.py Tue Mar 27 15:02:39 2007 +0900 @@ -580,9 +580,12 @@ class Guest(object): "networks": self._get_network_xml(install), \ "graphics": self._get_graphics_xml(install) } - def get_config_xml(self, install = True): + def get_config_xml(self, install = True, disk_boot = False): if install: - osblob = self._get_install_xml() + if disk_boot: + osblob = self._get_runtime_xml() + else: + osblob = self._get_install_xml() action = "destroy" else: osblob = self._get_runtime_xml() diff -r 36a9973c2e28 virtinst/FullVirtGuest.py --- a/virtinst/FullVirtGuest.py Tue Mar 27 08:13:38 2007 -0400 +++ b/virtinst/FullVirtGuest.py Wed Mar 28 16:29:01 2007 +0900 @@ -17,26 +17,27 @@ import Guest import Guest import util import DistroManager +import logging class FullVirtGuest(Guest.XenGuest): - OS_TYPES = { "Linux" : { "Red Hat Enterprise Linux AS 2.1/3" : { "acpi" : True, "apic": True }, \ - "Red Hat Enterprise Linux 4" : { "acpi" : True, "apic": True }, \ - "Red Hat Enterprise Linux 5" : { "acpi" : True, "apic": True }, \ - "Fedora Core 4-6" : { "acpi" : True, "apic": True }, \ - "Suse Linux Enterprise Server" : { "acpi" : True, "apic": True }, \ - "Other Linux 2.6 kernel" : { "acpi" : True, "apic": True } }, \ - "Microsoft Windows" : { "Windows 2000" : { "acpi": False, "apic" : False }, \ - "Windows XP" : { "acpi": True, "apic" : True }, \ - "Windows Server 2003" : { "acpi": True, "apic" : True }, \ - "Windows Vista" : { "acpi": True, "apic" : True } }, \ - "Novell Netware" : { "Netware 4" : { "acpi": True, "apic": True }, \ - "Netware 5" : { "acpi": True, "apic": True }, \ - "Netware 6" : { "acpi": True, "apic": True } }, \ - "Sun Solaris" : { "Solaris 10" : { "acpi": True, "apic": True }, \ - "Solaris 9" : { "acpi": True, "apic": True } }, \ - "Other" : { "MS-DOS" : { "acpi": False, "apic" : False }, \ - "Free BSD" : { "acpi": True, "apic" : True }, \ - "Other" : { "acpi": True, "apic" : True } } } + OS_TYPES = { "Linux" : { "Red Hat Enterprise Linux AS 2.1/3" : { "acpi" : True, "apic": True, "continue": False }, \ + "Red Hat Enterprise Linux 4" : { "acpi" : True, "apic": True, "continue": False }, \ + "Red Hat Enterprise Linux 5" : { "acpi" : True, "apic": True, "continue": False }, \ + "Fedora Core 4-6" : { "acpi" : True, "apic": True, "continue": False }, \ + "Suse Linux Enterprise Server" : { "acpi" : True, "apic": True, "continue": False }, \ + "Other Linux 2.6 kernel" : { "acpi" : True, "apic": True, "continue": False } }, \ + "Microsoft Windows" : { "Windows 2000" : { "acpi": False, "apic" : False, "continue": True }, \ + "Windows XP" : { "acpi": True, "apic" : True, "continue": True }, \ + "Windows Server 2003" : { "acpi": True, "apic" : True, "continue": True }, \ + "Windows Vista" : { "acpi": True, "apic" : True, "continue": True } }, \ + "Novell Netware" : { "Netware 4" : { "acpi": True, "apic": True, "continue": False }, \ + "Netware 5" : { "acpi": True, "apic": True, "continue": False }, \ + "Netware 6" : { "acpi": True, "apic": True, "continue": False } }, \ + "Sun Solaris" : { "Solaris 10" : { "acpi": True, "apic": True, "continue": False }, \ + "Solaris 9" : { "acpi": True, "apic": True, "continue": False } }, \ + "Other" : { "MS-DOS" : { "acpi": False, "apic" : False, "continue": False }, \ + "Free BSD" : { "acpi": True, "apic" : True, "continue": False }, \ + "Other" : { "acpi": True, "apic" : True, "continue": False } } } def __init__(self, type=None, arch=None, connection=None, hypervisorURI=None, emulator=None): Guest.Guest.__init__(self, type=type, connection=connection, hypervisorURI=hypervisorURI) @@ -190,3 +191,26 @@ class FullVirtGuest(Guest.XenGuest): self.disks.append(Guest.VirtualDisk(cdrom, device=Guest.VirtualDisk.DEVICE_CDROM, readOnly=True, transient=True)) return tmpfiles + + def get_continue_inst(self): + return FullVirtGuest.OS_TYPES[self._os_type][self._os_variant]["continue"] + + def continue_install(self, consolecb, meter): + install_xml = self.get_config_xml(disk_boot = True) + logging.debug("Starting guest from '%s'" % ( install_xml )) + meter.start(size=None, text="Starting domain...") + self.domain = self.conn.createLinux(install_xml, 0) + if self.domain is None: + raise RuntimeError, "Unable to start domain for guest, aborting installation!" + meter.end(0) + + self.connect_console(consolecb) + + # ensure there's time for the domain to finish destroying if the + # install has finished or the guest crashed + if consolecb: + time.sleep(1) + + # This should always work, because it'll lookup a config file + # for inactive guest, or get the still running install.. + return self.conn.lookupByName(self.name) From veillard at redhat.com Wed Mar 28 08:52:23 2007 From: veillard at redhat.com (Daniel Veillard) Date: Wed, 28 Mar 2007 04:52:23 -0400 Subject: [et-mgmt-tools] virt-manager new host overview & resources window In-Reply-To: <200703280038.l2S0c0Fm012374@fjmscan503.ms.jp.fujitsu.com> References: <20070328000120.GG3126@redhat.com> <200703280038.l2S0c0Fm012374@fjmscan503.ms.jp.fujitsu.com> Message-ID: <20070328085223.GE23514@redhat.com> On Wed, Mar 28, 2007 at 09:37:30AM +0900, Atsushi SAKAI wrote: > Hi, Dan > > Is there any plan to show Scheduler information details? > (weight, cap etc) > I am plan to make scheduler API on libvirt. > or this issue should be discussed in libvirt? That's likely to be quite specific to Xen, but as Dan suggested let's discuss that on the libvir-list, Daniel -- Red Hat Virtualization group http://redhat.com/virtualization/ Daniel Veillard | virtualization library http://libvirt.org/ veillard at redhat.com | libxml GNOME XML XSLT toolkit http://xmlsoft.org/ http://veillard.com/ | Rpmfind RPM search engine http://rpmfind.net/ From takatom at jp.fujitsu.com Wed Mar 28 09:47:51 2007 From: takatom at jp.fujitsu.com (Takahashi Tomohiro) Date: Wed, 28 Mar 2007 18:47:51 +0900 Subject: [et-mgmt-tools] About a solution to operate the virt-managerinchangeset 436 Message-ID: <026201c7711e$2a94a250$ecb1220a@fj98229700> Hi,Rich Any progress about this issue ? It is critical for me. > I'm getting a very similar error at the moment, which I'm looking into. > > Rich. From takatom at jp.fujitsu.com Wed Mar 28 13:01:24 2007 From: takatom at jp.fujitsu.com (Takahashi Tomohiro) Date: Wed, 28 Mar 2007 22:01:24 +0900 Subject: [et-mgmt-tools] [PATCH] Check the making domain's disk sources for virt-manager Message-ID: <02ad01c77139$35988110$ecb1220a@fj98229700> Hi, The virt-manager specify the Domain's disk sources. But the virt-manager doesn't check that the disk sources are conflicted among the other domains. The attached patch resolve this issue in the following way: 1) Count that the making domian's disk src is already use the exists domain's disk source. 2) if disk src is already use, confirmation message. Disk "/dev/xxx" is already in use by another guest! Do you really want to use the disk ? 3) if select yes then leave disk source as it is, else(no) specify disk source again. Signed-off-by: Tomohiro Takahashi Thanks, Tomohiro Takahashi. diff -uNrp virt-manager--devel_442.orig/src/virtManager/create.py virt-manager--devel_442/src/virtManager/create.py --- virt-manager--devel_442.orig/src/virtManager/create.py 2007-03-28 13:42:41.000000000 +0900 +++ virt-manager--devel_442/src/virtManager/create.py 2007-03-28 20:58:52.000000000 +0900 @@ -57,6 +57,8 @@ PAGE_SUMMARY = 8 KEYBOARD_DIR = "/etc/sysconfig/keyboard" +BUTTON_YES = -8 + class vmmCreateMeter(progress.BaseMeter): def __init__(self, asyncjob): # progress meter has to run asynchronously, so pass in the @@ -817,6 +819,13 @@ class vmmCreate(gobject.GObject): self._validation_error_box(_("Storage Address Is Directory"), \ _("You chose 'Simple File' storage for your storage method, but chose a directory instead of a file. Please enter a new filename or choose an existing file.")) return False + + d = virtinst.VirtualDisk(self.get_config_disk_image(), self.get_config_disk_size(), sparse = self.is_sparse_file()) + if d.is_conflict_disk(self.connection.vmm) is True: + res = self._validation_error_box_YES_NO(_('Disk "%s" is already in use by another guest!' % disk), \ + _("Do you really want to use the disk ?")) + return res + elif page_num == PAGE_NETWORK: if self.window.get_widget("net-type-network").get_active(): if self.window.get_widget("net-network").get_active() == -1: @@ -844,6 +853,22 @@ class vmmCreate(gobject.GObject): message_box.format_secondary_text(text2) message_box.run() message_box.destroy() + + def _validation_error_box_YES_NO(self, text1, text2=None): + #import pdb; pdb.set_trace() + message_box = gtk.MessageDialog(self.window.get_widget("vmm-create"), \ + 0, \ + gtk.MESSAGE_WARNING, \ + gtk.BUTTONS_YES_NO, \ + text1) + if text2 != None: + message_box.format_secondary_text(text2) + if message_box.run()== BUTTON_YES: + res = True + else: + res = False + message_box.destroy() + return res def populate_opt_media(self, model): # get a list of optical devices with data discs in, for FV installs From mdehaan at redhat.com Wed Mar 28 15:14:37 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 28 Mar 2007 11:14:37 -0400 Subject: [et-mgmt-tools] Cobbler bugfix "--repos" parameter on "profile add" Message-ID: <460A865D.9070709@redhat.com> Thanks to Matt S. for finding this... For those wanting to use the repo management (connecting repository mirrors with profiles), this change, along with the fixes to cobbler_syslogd are now checked in to git. git clone git://et.redhat.com/cobbler make rpm -Uvh cobbler*.rpm This particular change was fairly simple, what happened was we're saving the repo list as an array (ideally) but it was being stored as a string. This caused bad things to happen. --- a/cobbler/item_profile.py Fri Mar 23 16:36:46 2007 -0400 +++ b/cobbler/item_profile.py Wed Mar 28 11:00:31 2007 -0400 @@ -95,7 +95,7 @@ class Profile(item.Item): ok = False break if ok: - self.repos = repos + self.repos = repolist else: raise cexceptions.CobblerException("no_repos") From hbrock at redhat.com Wed Mar 28 18:40:50 2007 From: hbrock at redhat.com (Hugh Brock) Date: Wed, 28 Mar 2007 14:40:50 -0400 Subject: [et-mgmt-tools] [PATCH] [RESEND]Check the making domain's mac address(part 2) In-Reply-To: <460A2ADE.5030007@aa.jp.fujitsu.com> References: <45FE5301.3020300@aa.jp.fujitsu.com> <4607AFD5.7030107@aa.jp.fujitsu.com> <4607F0A3.8040307@redhat.com> <46089CE2.5000802@aa.jp.fujitsu.com> <46093BB5.5080000@redhat.com> <460A2ADE.5030007@aa.jp.fujitsu.com> Message-ID: <460AB6B2.9070208@redhat.com> Tatsuro Enokura wrote: > Hi Hugh > > > I agree with your suggestion. > I rewrite a patch that the mac-address conflict checking code in > VirtualNetworkInterface.setup() so that it prints a warning to > the stderr and logs. > > Thanks, > Tatsuro Enokura > This is much better, I have applied it (after fixing a bug we introduced that was getting in the way of specifying a mac address at all). Thanks! Take care, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From matts at hostingsupport.com Wed Mar 28 19:00:29 2007 From: matts at hostingsupport.com (Matt S [UNIX Administrator]) Date: Wed, 28 Mar 2007 14:00:29 -0500 Subject: [et-mgmt-tools] Cobbler bugfix "--repos" parameter on "profile add" In-Reply-To: <460A865D.9070709@redhat.com> References: <460A865D.9070709@redhat.com> Message-ID: <1175108429.5768.39.camel@shire.propagation.net> My git clone attempt fails. git clone git://et.redhat.com/cobbler Initialized empty Git repository in /root/cobbler/.git/ fatal: unable to connect a socket (Connection timed out) On Wed, 2007-03-28 at 11:14 -0400, Michael DeHaan wrote: > Thanks to Matt S. for finding this... > > For those wanting to use the repo management (connecting repository > mirrors with profiles), this change, along with the fixes to > cobbler_syslogd are now checked in to git. > > git clone git://et.redhat.com/cobbler > make > rpm -Uvh cobbler*.rpm > > This particular change was fairly simple, what happened was we're saving > the repo list as an array (ideally) but it was being stored as a > string. This caused bad things to happen. > > --- a/cobbler/item_profile.py Fri Mar 23 16:36:46 2007 -0400 > +++ b/cobbler/item_profile.py Wed Mar 28 11:00:31 2007 -0400 > @@ -95,7 +95,7 @@ class Profile(item.Item): > ok = False > break > if ok: > - self.repos = repos > + self.repos = repolist > else: > raise cexceptions.CobblerException("no_repos") > > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From hbrock at redhat.com Wed Mar 28 19:28:47 2007 From: hbrock at redhat.com (Hugh Brock) Date: Wed, 28 Mar 2007 15:28:47 -0400 Subject: [et-mgmt-tools] [PATCH] Check the making domain's disk sources for virt-manager In-Reply-To: <02ad01c77139$35988110$ecb1220a@fj98229700> References: <02ad01c77139$35988110$ecb1220a@fj98229700> Message-ID: <460AC1EF.5090709@redhat.com> Takahashi Tomohiro wrote: > Hi, > > The virt-manager specify the Domain's disk sources. > But the virt-manager doesn't check that the disk sources are > conflicted among the other domains. > > The attached patch resolve this issue in the following way: > > 1) Count that the making domian's disk src is already use > the exists domain's disk source. > 2) if disk src is already use, confirmation message. > Disk "/dev/xxx" is already in use by another guest! > Do you really want to use the disk ? > 3) if select yes then leave disk source as it is, > else(no) specify disk source again. > > Signed-off-by: Tomohiro Takahashi > > Thanks, > Tomohiro Takahashi. > I have applied this with a minor change: you can use gtk.RESPONSE_YES for the response constant, rather than creating your own. It would be nice, however, to tell the user which guest is using the disk? Maybe another patch that adds that capability would be good. Thanks, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From mdehaan at redhat.com Wed Mar 28 19:59:33 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 28 Mar 2007 15:59:33 -0400 Subject: git port (was Re: [et-mgmt-tools] Cobbler bugfix "--repos" parameter on "profile add") In-Reply-To: <1175108429.5768.39.camel@shire.propagation.net> References: <460A865D.9070709@redhat.com> <1175108429.5768.39.camel@shire.propagation.net> Message-ID: <460AC925.7040909@redhat.com> Matt S [UNIX Administrator] wrote: > My git clone attempt fails. > > git clone git://et.redhat.com/cobbler > Initialized empty Git repository in /root/cobbler/.git/ > fatal: unable to connect a socket (Connection timed out) > There is a chance there is a problem on our end, but as this is an external server (even to me), I wouldn't think this would be the case. Possibly a firewall is in the way? The git port is 9418. There is a way to pull git content over http:// that is less efficient, though I'm not sure if that is enabled. I'm checking on it. From hbrock at redhat.com Wed Mar 28 21:10:42 2007 From: hbrock at redhat.com (Hugh Brock) Date: Wed, 28 Mar 2007 17:10:42 -0400 Subject: [et-mgmt-tools] Subject:[PATCH][RESEND] Windows HVM can't continue toinstall In-Reply-To: <200703281750.IGC73478.O9G80HK4@aa.jp.fujitsu.com> References: <200703201857.AEF73948.08KGO94H@aa.jp.fujitsu.com> <4602DC54.7050203@redhat.com> <200703281750.IGC73478.O9G80HK4@aa.jp.fujitsu.com> Message-ID: <460AD9D2.7060407@redhat.com> Nobuhiro Itou wrote: > Hi Hugh > >> We don't really want to put OS-specific stuff into the virt-install code >> if we can help it; we made an exception for the config flag setting >> stuff because we're just abstracting behavior (setting acpi= and apic=) >> that is already available. > > All right. > In fact, I think OS-specific stuff shouldn't put into the virt-install, too. > >> However I don't see any reason why a patch that would allow a virtinst >> argument like "--keep-cdrom" and then put something like "> type="block" device="cdrom">... etc" in the domain's permanent xml >> wouldn't work. This would mean that the domain would have access to the >> cdrom drive every time it reboots. Of course, it also means we'll need >> to provide some way to disconnect the domain from that drive once the >> install process is finished, but we'll get to that later. >> >> Let me know if that works for you... > > I remaked a patch. > This time, it has the flag in FullVirtGuest.OS_TYPES. > The OS-specific code is removed from virt-inst. > How about this correction? > > Signed-off-by: Nobuhiro Itou > > > Thanks, > Nobuhiro Itou. > OK, I have applied this. It seems to work fine with Win2k, but I don't have XP install CDs, so I can't tell if it works or not with XP or Server 2003. Please test it and let me know! And Thanks!, as always... Take care, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From fj0873gn at aa.jp.fujitsu.com Thu Mar 29 02:41:17 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Thu, 29 Mar 2007 11:41:17 +0900 Subject: [et-mgmt-tools] Subject:[PATCH][RESEND] Windows HVM can't continuetoinstall In-Reply-To: <460AD9D2.7060407@redhat.com> References: <200703201857.AEF73948.08KGO94H@aa.jp.fujitsu.com> <4602DC54.7050203@redhat.com> <200703281750.IGC73478.O9G80HK4@aa.jp.fujitsu.com> <460AD9D2.7060407@redhat.com> Message-ID: <200703291141.IEH87566.0OK8HG49@aa.jp.fujitsu.com> Hi Hugh > OK, I have applied this. It seems to work fine with Win2k, but I don't > have XP install CDs, so I can't tell if it works or not with XP or > Server 2003. Please test it and let me know! I tested Windows Server 2003. It works fine with Server 2003, too. But, I also don't have XP install CDs. As for XP, if there is an opportunity, I'll try it. Thanks, Nobuhiro Itou. From fj7716hz at aa.jp.fujitsu.com Thu Mar 29 09:16:26 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Thu, 29 Mar 2007 18:16:26 +0900 Subject: [et-mgmt-tools] [PATCH] HVM 3rd disk name don't use "hdc" In-Reply-To: <460718BB.7040602@aa.jp.fujitsu.com> References: <4602673D.8060802@aa.jp.fujitsu.com> <20070322152616.GD22912@redhat.com> <460718BB.7040602@aa.jp.fujitsu.com> Message-ID: <460B83EA.4080305@aa.jp.fujitsu.com> Hi, Dan Tatsuro Enokura wrote: >>> I try that making the HVM domain with 3+ disks by the virt-install command, >>> but the command says following error message: >>> # virt-install --name=HVM_3DISKS --file=/dev/sda6 --file=/dev/sda7 \ >>> --file=/dev/sda8 --hvm --vnc --ram=376 --cdrom=/dev/cdrom >> This is a ticky one ! We need to a) refuse to allow > 3 --file arguments >> to be used with HVM guests, and b) make sure the 3rd file gets given >> hdd as its device name. > > Thank you for your reviewing. > I rewite the patch. Would you give me a comment on this patch? If not, please apply it. Thanks, Tatsuro Enokura -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: diskcheck.patch URL: From fj0873gn at aa.jp.fujitsu.com Thu Mar 29 11:41:55 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Thu, 29 Mar 2007 20:41:55 +0900 Subject: [et-mgmt-tools] Subject:[PATCH][RESEND] Windows HVM can't continue to install In-Reply-To: <460AD9D2.7060407@redhat.com> References: <200703201857.AEF73948.08KGO94H@aa.jp.fujitsu.com> <4602DC54.7050203@redhat.com> <200703281750.IGC73478.O9G80HK4@aa.jp.fujitsu.com> <460AD9D2.7060407@redhat.com> Message-ID: <200703292041.DHD90640.8HO09KG4@aa.jp.fujitsu.com> Hi Hugh > I tested Windows Server 2003. It works fine with Server 2003, too. > But, I also don't have XP install CDs. > As for XP, if there is an opportunity, I'll try it. I made a mistake. The time module was not imported. please apply this correction. Thanks, Nobuhiro Itou. diff -r 0a4efe737df9 virtinst/FullVirtGuest.py --- a/virtinst/FullVirtGuest.py Wed Mar 28 17:01:09 2007 -0400 +++ b/virtinst/FullVirtGuest.py Thu Mar 29 19:45:42 2007 +0900 @@ -18,6 +18,7 @@ import util import util import DistroManager import logging +import time class FullVirtGuest(Guest.XenGuest): OS_TYPES = { "Linux" : { "Red Hat Enterprise Linux AS 2.1/3" : { "acpi" : True, "apic": True, "continue": False }, \ From hbrock at redhat.com Thu Mar 29 19:48:49 2007 From: hbrock at redhat.com (Hugh Brock) Date: Thu, 29 Mar 2007 15:48:49 -0400 Subject: [et-mgmt-tools] [PATCH] HVM 3rd disk name don't use "hdc" In-Reply-To: <460B83EA.4080305@aa.jp.fujitsu.com> References: <4602673D.8060802@aa.jp.fujitsu.com> <20070322152616.GD22912@redhat.com> <460718BB.7040602@aa.jp.fujitsu.com> <460B83EA.4080305@aa.jp.fujitsu.com> Message-ID: <460C1821.3040707@redhat.com> Tatsuro Enokura wrote: > Hi, Dan > > Tatsuro Enokura wrote: >>>> I try that making the HVM domain with 3+ disks by the virt-install command, >>>> but the command says following error message: >>>> # virt-install --name=HVM_3DISKS --file=/dev/sda6 --file=/dev/sda7 \ >>>> --file=/dev/sda8 --hvm --vnc --ram=376 --cdrom=/dev/cdrom >>> This is a ticky one ! We need to a) refuse to allow > 3 --file arguments >>> to be used with HVM guests, and b) make sure the 3rd file gets given >>> hdd as its device name. >> Thank you for your reviewing. >> I rewite the patch. > > Would you give me a comment on this patch? > If not, please apply it. > > Thanks, > Tatsuro Enokura > Yes, this looks fine, I have applied it. Thanks!! --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From mdehaan at redhat.com Thu Mar 29 22:22:14 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 29 Mar 2007 18:22:14 -0400 Subject: git port (was Re: [et-mgmt-tools] Cobbler bugfix "--repos" parameter on "profile add") In-Reply-To: <460AC925.7040909@redhat.com> References: <460A865D.9070709@redhat.com> <1175108429.5768.39.camel@shire.propagation.net> <460AC925.7040909@redhat.com> Message-ID: <460C3C16.90109@redhat.com> Michael DeHaan wrote: > Matt S [UNIX Administrator] wrote: >> My git clone attempt fails. >> git clone git://et.redhat.com/cobbler >> Initialized empty Git repository in /root/cobbler/.git/ >> fatal: unable to connect a socket (Connection timed out) >> > > There is a chance there is a problem on our end, but as this is an > external server (even to me), I wouldn't think this would be the case. > > Possibly a firewall is in the way? The git port is 9418. > > There is a way to pull git content over http:// that is less > efficient, though I'm not sure if that is enabled. I'm checking on it. > > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools the git:// URLs should work now. If they don't work for you, let us know. --Michael From matts at hostingsupport.com Thu Mar 29 22:27:28 2007 From: matts at hostingsupport.com (Matt S [UNIX Administrator]) Date: Thu, 29 Mar 2007 17:27:28 -0500 Subject: git port (was Re: [et-mgmt-tools] Cobbler bugfix "--repos" parameter on "profile add") In-Reply-To: <460C3C16.90109@redhat.com> References: <460A865D.9070709@redhat.com> <1175108429.5768.39.camel@shire.propagation.net> <460AC925.7040909@redhat.com> <460C3C16.90109@redhat.com> Message-ID: <1175207248.5768.41.camel@shire.propagation.net> Thanks, it's working now. On Thu, 2007-03-29 at 18:22 -0400, Michael DeHaan wrote: > Michael DeHaan wrote: > > Matt S [UNIX Administrator] wrote: > >> My git clone attempt fails. > >> git clone git://et.redhat.com/cobbler > >> Initialized empty Git repository in /root/cobbler/.git/ > >> fatal: unable to connect a socket (Connection timed out) > >> > > > > There is a chance there is a problem on our end, but as this is an > > external server (even to me), I wouldn't think this would be the case. > > > > Possibly a firewall is in the way? The git port is 9418. > > > > There is a way to pull git content over http:// that is less > > efficient, though I'm not sure if that is enabled. I'm checking on it. > > > > > > > > _______________________________________________ > > et-mgmt-tools mailing list > > et-mgmt-tools at redhat.com > > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > the git:// URLs should work now. If they don't work for you, let us know. > > --Michael > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From fj0588di at aa.jp.fujitsu.com Fri Mar 30 08:16:24 2007 From: fj0588di at aa.jp.fujitsu.com (S.Sakamoto) Date: Fri, 30 Mar 2007 17:16:24 +0900 Subject: [et-mgmt-tools] [PATCH] check a UUID format Message-ID: <200703301716.FHI87074.9GK0JE69@aa.jp.fujitsu.com> Hi When I install by virt-intall with invalid UUID(e.g."--uuid=4096 Characters"), virt-install freeze. So, here's the patch adds to check a UUID with following format: xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx ( xx should be in hexadecimal ) Signed-off-by: Shigeki Sakamoto Thanks, Shigeki Sakamoto. ========================================================== diff -r e63d9c7f0a89 virt-install --- a/virt-install Thu Mar 29 15:56:42 2007 -0400 +++ b/virt-install Fri Mar 30 17:07:23 2007 +0900 @@ -88,6 +88,7 @@ def get_uuid(uuid, guest): guest.uuid = uuid except ValueError, e: print "ERROR: ", e + sys.exit(1) def get_vcpus(vcpus, check_cpu, guest, conn): while 1: @@ -305,7 +306,7 @@ def parse_args(): help="Memory to allocate for guest instance in megabytes") parser.add_option("-u", "--uuid", type="string", dest="uuid", action="callback", callback=check_before_store, - help="UUID for the guest; if none is given a random UUID will be generated") + help="UUID for the guest; if none is given a random UUID will be generated. if specify UUID, you should use hexadecimal number with a form of XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX.") parser.add_option("", "--vcpus", type="int", dest="vcpus", help="Number of vcpus to configure for your guest") parser.add_option("", "--check-cpu", action="store_true", dest="check_cpu", diff -r e63d9c7f0a89 virtinst/Guest.py --- a/virtinst/Guest.py Thu Mar 29 15:56:42 2007 -0400 +++ b/virtinst/Guest.py Fri Mar 30 17:04:39 2007 +0900 @@ -417,12 +417,10 @@ class Guest(object): return self._uuid def set_uuid(self, val): # need better validation - if type(val) == type("str"): - self._uuid = val - elif type(val) == type(123): - self._uuid = util.uuidToString(val) - else: - raise ValueError, "Invalid value for UUID" + form = re.match("[a-fA-F0-9]{8}[-]([a-fA-F0-9]{4}[-]){3}[a-fA-F0-9]{12}", val) + if form is None: + raise ValueError, "UUID must use hexadecimal number with a form of XXXXXXXX-XXXX-XXXX-XXXX-XXXXXXXXXXXX" + self._uuid = val uuid = property(get_uuid, set_uuid) From satyaakam at gmail.com Fri Mar 30 13:23:53 2007 From: satyaakam at gmail.com (satyaakam goswami) Date: Fri, 30 Mar 2007 18:53:53 +0530 Subject: [et-mgmt-tools] RHEL 3.0 Update 5 Hangs Message-ID: <6491e1350703300623s52f55a7ey22ba1e799f0da37d@mail.gmail.com> Hi, I am trying full virt install as DomU for RHEL 3 Update 5 using virt-manager , after booting the firt CD the VM hangs with the following message CALLIN, before setup_local_APIC() Calibrating delay loop... i tried giving noacpi kernel boot option it did not help attached is the screen shot Satya -------------- next part -------------- A non-text attachment was scrubbed... Name: console.png Type: image/png Size: 8898 bytes Desc: not available URL: From jeevanullas at gmail.com Fri Mar 30 13:51:37 2007 From: jeevanullas at gmail.com (Deependra Shekhawat) Date: Fri, 30 Mar 2007 19:21:37 +0530 Subject: [et-mgmt-tools] Want to install EL5 via xen. Message-ID: <57127b5f0703300651i1d978024k46930a3c6bd17531@mail.gmail.com> Hello guys, I have installed EL5 via xen. Now I wanted to try cobbler + koan to install EL5. Somewhat like un-attended virtual machine installation. I have installed cobbler and koan from the python code. I did 'cobbler check' and it only said about /etc/dhcpd.conf. It says to include next-servers line. But why I need it when I have DHCP on the same server as I have xen, cobbler and koan. Other things are setup well. But now I don't know how do I proceed. If someone have been through this thing please help. Regards Deependra -- Enjoy Life ! -------------- next part -------------- An HTML attachment was scrubbed... URL: From dhawal at netmagicsolutions.com Fri Mar 30 13:59:22 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Fri, 30 Mar 2007 19:29:22 +0530 Subject: [et-mgmt-tools] Want to install EL5 via xen. In-Reply-To: <57127b5f0703300651i1d978024k46930a3c6bd17531@mail.gmail.com> References: <57127b5f0703300651i1d978024k46930a3c6bd17531@mail.gmail.com> Message-ID: <460D17BA.6070201@netmagicsolutions.com> Deependra Shekhawat wrote: > Hello guys, > > I have installed EL5 via xen. Now I wanted to try cobbler + koan to > install EL5. Somewhat like un-attended virtual machine installation. I > have installed cobbler and koan from the python code. I did 'cobbler > check' and it only said about /etc/dhcpd.conf. It says to include > next-servers line. But why I need it when I have DHCP on the same server > as I have xen, cobbler and koan. Other things are setup well. > But now I don't know how do I proceed. If someone have been through this > thing please help. From 'man dhcpd.conf' The next-server statement: The next-server statement is used to specify the host address of the server from which the initial boot file (specified in the filename statement) is to be loaded. Server-name should be a numeric IP address or a domain name. If no next-server parameter applies to a given client, the DHCP server's IP address is used. To summarize, you could leave it null if your dhcp/pxe server are the same, else mention the IP address of the pxe server. If cobbler insists on the next-server line, then try using your local IP address. From jeevanullas at gmail.com Fri Mar 30 14:05:08 2007 From: jeevanullas at gmail.com (Deependra Shekhawat) Date: Fri, 30 Mar 2007 19:35:08 +0530 Subject: [et-mgmt-tools] Want to install EL5 via xen. In-Reply-To: <460D17BA.6070201@netmagicsolutions.com> References: <57127b5f0703300651i1d978024k46930a3c6bd17531@mail.gmail.com> <460D17BA.6070201@netmagicsolutions.com> Message-ID: <57127b5f0703300705o694325f2h5c2b0fac625388eb@mail.gmail.com> Thanks for the quick response. I am moving forward with my learning. On 3/30/07, Dhawal Doshy wrote: > > Deependra Shekhawat wrote: > > Hello guys, > > > > I have installed EL5 via xen. Now I wanted to try cobbler + koan to > > install EL5. Somewhat like un-attended virtual machine installation. I > > have installed cobbler and koan from the python code. I did 'cobbler > > check' and it only said about /etc/dhcpd.conf. It says to include > > next-servers line. But why I need it when I have DHCP on the same server > > as I have xen, cobbler and koan. Other things are setup well. > > But now I don't know how do I proceed. If someone have been through this > > thing please help. > > From 'man dhcpd.conf' > The next-server statement: > The next-server statement is used to specify the host address of the > server from which the initial boot file (specified in the filename > statement) is to be loaded. Server-name should be a numeric IP address > or a domain name. If no next-server parameter applies to a given client, > the DHCP server's IP address is used. > > To summarize, you could leave it null if your dhcp/pxe server are the > same, else mention the IP address of the pxe server. If cobbler insists > on the next-server line, then try using your local IP address. > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > -- Enjoy Life ! -------------- next part -------------- An HTML attachment was scrubbed... URL: From John.Sanabria at ece.uprm.edu Fri Mar 30 15:27:27 2007 From: John.Sanabria at ece.uprm.edu (John Sanabria) Date: Fri, 30 Mar 2007 11:27:27 -0400 (AST) Subject: [et-mgmt-tools] what happen? :( Message-ID: <50745.136.145.116.114.1175268447.squirrel@ece.uprm.edu> Hi, the last week i had executed the next line command to create a virtual nodes in 136.145.116.76 host: cobbler enchant --virt=yes --address=136.145.116.76 --profile="xentemplate-profile" and it work. Today, i try again to execute the command and appear that message: running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias nocobblerhostkey' root at 136.145.116.76 wget http://136.145.116.114/cobbler/koan-0.2.8-1.noarch.rpm Warning: Permanently added 'nocobblerhostkey' (RSA) to the list of known hosts. --11:18:07-- http://136.145.116.114/cobbler/koan-0.2.8-1.noarch.rpm Connecting to 136.145.116.114:80... connected. HTTP request sent, awaiting response... 200 OK Length: 93410 (91K) [application/x-rpm] Saving to: `koan-0.2.8-1.noarch.rpm' 0K .......... .......... .......... .......... .......... 54% 4.08M 0s 50K .......... .......... .......... .......... . 100% 14.1M=0.01s 11:18:08 (6.00 MB/s) - `koan-0.2.8-1.noarch.rpm' saved [93410/93410] returns: 0 running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias nocobblerhostkey' root at 136.145.116.76 up2date install syslinux libvirt-python libvirt bash: up2date: command not found returns: 127 running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias nocobblerhostkey' root at 136.145.116.76 yum -y install syslinux libvirt-python libvirt Loading "installonlyn" plugin Setting up Install Process Setting up repositories Reading repository metadata in from local files Parsing package install arguments Nothing to do returns: 0 running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias nocobblerhostkey' root at 136.145.116.76 rpm -Uvh koan-0.2.8-1.noarch.rpm --force --nodeps Preparing... ################################################## koan ################################################## returns: 0 running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias nocobblerhostkey' root at 136.145.116.76 koan --virt --profile=xentemplate-profile --server=136.145.116.114 - processing profile: xentemplate-profile - fetching configuration for profile: xentemplate-profile - url=http://136.145.116.114/cobbler/profiles/xentemplate-profile - {'kickstart': 'http://136.145.116.114/cobbler_track/kickstarts/xentemplate-profile/ks.cfg', 'name': 'xentemplate-profile', 'virt_ram': 256, 'repos': '', 'kernel_options': 'lang= ksdevice=eth0 text syslog=136.145.116.114:25150 devfs=nomount ramdisk_size=16438 append ', 'virt_name': 'xentemplate', 'virt_file_size': 6, 'distro': 'xentemplate', 'virt_paravirt': 'True', 'ks_meta': ''} - fetching configuration for distro: xentemplate - url=http://136.145.116.114/cobbler/distros/xentemplate - {'kernel': '/root/vmlinuz', 'ks_meta': '', 'breed': 'redhat', 'kernel_options': 'lang= ksdevice=eth0 text syslog=136.145.116.114:25150 devfs=nomount ramdisk_size=16438 append ', 'initrd': '/root/initrd.img', 'arch': 'x86', 'name': 'xentemplate'} - downloading initrd initrd.img to /tmp/initrd.img - url=http://136.145.116.114/cobbler/images/xentemplate/initrd.img -libvir: error : no support for hypervisor Traceback (most recent call last): File "/usr/lib/python2.4/site-packages/koan/app.py", line 97, in main k.run() File "/usr/lib/python2.4/site-packages/koan/app.py", line 152, in run self.do_virt() File "/usr/lib/python2.4/site-packages/koan/app.py", line 297, in do_virt return self.do_net_install("/tmp",after_download) File "/usr/lib/python2.4/site-packages/koan/app.py", line 253, in do_net_install after_download(self, distro_data, profile_data) File "/usr/lib/python2.4/site-packages/koan/app.py", line 296, in after_download self.do_virt_net_install(profile_data, distro_data) File "/usr/lib/python2.4/site-packages/koan/app.py", line 600, in do_virt_net_install extra=kextra File "/usr/lib/python2.4/site-packages/koan/virtcreate.py", line 84, in start_paravirt_install guest = virtinst.ParaVirtGuest() File "/usr/lib/python2.4/site-packages/virtinst/ParaVirtGuest.py", line 37, in __init__ XenGuest.XenGuest.__init__(self, hypervisorURI=hypervisorURI) File "/usr/lib/python2.4/site-packages/virtinst/XenGuest.py", line 183, in __init__ self.conn = libvirt.open(hypervisorURI) downloading kernel vmlinuz to /tmp/vmlinuz - url=http://136.145.116.114/cobbler/images/xentemplate/vmlinuz - kernel saved = /tmp/vmlinuz - initrd saved = /tmp/initrd.img File "/usr/lib/python2.4/site-packages/libvirt.py", line 95, in open if ret is None:raise libvirtError('virConnectOpen() failed') libvirtError: virConnectOpen() failed no support for hypervisor returns: 3 Traceback (most recent call last): File "/usr/lib/python2.4/site-packages/cobbler/action_enchant.py", line 101, in run self.ssh_exec("koan %s --profile=%s --server=%s" % (operation, self.profile, self.settings.server)) File "/usr/lib/python2.4/site-packages/cobbler/action_enchant.py", line 57, in ssh_exec raise cexceptions.CobblerException("enchant_failed","ssh failure") CobblerException: 'enchant failed (ssh failure)' enchant failed (exception) ---------------------------- Any suggestions. the machine where command were submitted has the next configuration: uname -a : "Linux reboot 2.6.19-1.2911.6.5.fc6xen #1 SMP Sun Mar 4 16:59:41 EST 2007 i686 i686 i386 GNU/Linux" koan: 0.2.7-1.fc6 cobbler: 0.4.3-0.fc6 thanks a lot for your answer. http://ece.uprm.edu/~s047267 http://del.icio.us/josanabr http://blog-grid.blogspot.com From mdehaan at redhat.com Fri Mar 30 17:22:40 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 30 Mar 2007 13:22:40 -0400 Subject: [et-mgmt-tools] what happen? :( In-Reply-To: <50745.136.145.116.114.1175268447.squirrel@ece.uprm.edu> References: <50745.136.145.116.114.1175268447.squirrel@ece.uprm.edu> Message-ID: <460D4760.8070002@redhat.com> John Sanabria wrote: > Hi, > > the last week i had executed the next line command to create a virtual > nodes in 136.145.116.76 host: > > cobbler enchant --virt=yes --address=136.145.116.76 > --profile="xentemplate-profile" > > and it work. > > Today, i try again to execute the command and appear that message: > > running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias > nocobblerhostkey' root at 136.145.116.76 wget > http://136.145.116.114/cobbler/koan-0.2.8-1.noarch.rpm > Warning: Permanently added 'nocobblerhostkey' (RSA) to the list of known > hosts. > --11:18:07-- http://136.145.116.114/cobbler/koan-0.2.8-1.noarch.rpm > Connecting to 136.145.116.114:80... connected. > HTTP request sent, awaiting response... 200 OK > Length: 93410 (91K) [application/x-rpm] > Saving to: `koan-0.2.8-1.noarch.rpm' > > 0K .......... .......... .......... .......... .......... 54% 4.08M 0s > 50K .......... .......... .......... .......... . 100% > 14.1M=0.01s > > 11:18:08 (6.00 MB/s) - `koan-0.2.8-1.noarch.rpm' saved [93410/93410] > > returns: 0 > running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias > nocobblerhostkey' root at 136.145.116.76 up2date install syslinux > libvirt-python libvirt > bash: up2date: command not found > returns: 127 > running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias > nocobblerhostkey' root at 136.145.116.76 yum -y install syslinux > libvirt-python libvirt > Loading "installonlyn" plugin > Setting up Install Process > Setting up repositories > Reading repository metadata in from local files > Parsing package install arguments > Nothing to do > returns: 0 > running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias > nocobblerhostkey' root at 136.145.116.76 rpm -Uvh koan-0.2.8-1.noarch.rpm > --force --nodeps > Preparing... > ################################################## > koan > ################################################## > returns: 0 > running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias > nocobblerhostkey' root at 136.145.116.76 koan --virt > --profile=xentemplate-profile --server=136.145.116.114 > - processing profile: xentemplate-profile > - fetching configuration for profile: xentemplate-profile > - url=http://136.145.116.114/cobbler/profiles/xentemplate-profile > - {'kickstart': > 'http://136.145.116.114/cobbler_track/kickstarts/xentemplate-profile/ks.cfg', > 'name': 'xentemplate-profile', 'virt_ram': 256, 'repos': '', > 'kernel_options': 'lang= ksdevice=eth0 text syslog=136.145.116.114:25150 > devfs=nomount ramdisk_size=16438 append ', 'virt_name': 'xentemplate', > 'virt_file_size': 6, 'distro': 'xentemplate', 'virt_paravirt': 'True', > 'ks_meta': ''} > - fetching configuration for distro: xentemplate > - url=http://136.145.116.114/cobbler/distros/xentemplate > - {'kernel': '/root/vmlinuz', 'ks_meta': '', 'breed': 'redhat', > 'kernel_options': 'lang= ksdevice=eth0 text syslog=136.145.116.114:25150 > devfs=nomount ramdisk_size=16438 append ', 'initrd': '/root/initrd.img', > 'arch': 'x86', 'name': 'xentemplate'} > - downloading initrd initrd.img to /tmp/initrd.img > - url=http://136.145.116.114/cobbler/images/xentemplate/initrd.img > -libvir: error : no support for hypervisor > Traceback (most recent call last): > File "/usr/lib/python2.4/site-packages/koan/app.py", line 97, in main > k.run() > File "/usr/lib/python2.4/site-packages/koan/app.py", line 152, in run > self.do_virt() > File "/usr/lib/python2.4/site-packages/koan/app.py", line 297, in do_virt > return self.do_net_install("/tmp",after_download) > File "/usr/lib/python2.4/site-packages/koan/app.py", line 253, in > do_net_install > after_download(self, distro_data, profile_data) > File "/usr/lib/python2.4/site-packages/koan/app.py", line 296, in > after_download > self.do_virt_net_install(profile_data, distro_data) > File "/usr/lib/python2.4/site-packages/koan/app.py", line 600, in > do_virt_net_install > extra=kextra > File "/usr/lib/python2.4/site-packages/koan/virtcreate.py", line 84, in > start_paravirt_install > guest = virtinst.ParaVirtGuest() > File "/usr/lib/python2.4/site-packages/virtinst/ParaVirtGuest.py", line > 37, in __init__ > XenGuest.XenGuest.__init__(self, hypervisorURI=hypervisorURI) > File "/usr/lib/python2.4/site-packages/virtinst/XenGuest.py", line 183, > in __init__ > self.conn = libvirt.open(hypervisorURI) > downloading kernel vmlinuz to /tmp/vmlinuz > - url=http://136.145.116.114/cobbler/images/xentemplate/vmlinuz > - kernel saved = /tmp/vmlinuz > - initrd saved = /tmp/initrd.img > File "/usr/lib/python2.4/site-packages/libvirt.py", line 95, in open > if ret is None:raise libvirtError('virConnectOpen() failed') > libvirtError: virConnectOpen() failed no support for hypervisor > returns: 3 > Traceback (most recent call last): > File "/usr/lib/python2.4/site-packages/cobbler/action_enchant.py", line > 101, in run > self.ssh_exec("koan %s --profile=%s --server=%s" % (operation, > self.profile, self.settings.server)) > File "/usr/lib/python2.4/site-packages/cobbler/action_enchant.py", line > 57, in ssh_exec > raise cexceptions.CobblerException("enchant_failed","ssh failure") > CobblerException: 'enchant failed (ssh failure)' > enchant failed (exception) > > ---------------------------- > > Any suggestions. > > the machine where command were submitted has the next configuration: > > uname -a : "Linux reboot 2.6.19-1.2911.6.5.fc6xen #1 SMP Sun Mar 4 > 16:59:41 EST 2007 i686 i686 i386 GNU/Linux" > koan: 0.2.7-1.fc6 > cobbler: 0.4.3-0.fc6 > > thanks a lot for your answer. > > > > > http://ece.uprm.edu/~s047267 > http://del.icio.us/josanabr > http://blog-grid.blogspot.com > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > I think the key is... "-libvir: error : no support for hypervisor" From the above, it looks like you're getting an error from Xen or libvirt, rather than from something cobbler/koan related. Are you sure you are installing a Xen kernel image vs a plain one? Is xend running? I'm really not super-up on libvirt error messages, so perhaps Dan/Hugh can field this. You could always ask fedora-xen as well. --Michael From mdehaan at redhat.com Fri Mar 30 17:26:27 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 30 Mar 2007 13:26:27 -0400 Subject: [et-mgmt-tools] Want to install EL5 via xen. In-Reply-To: <57127b5f0703300705o694325f2h5c2b0fac625388eb@mail.gmail.com> References: <57127b5f0703300651i1d978024k46930a3c6bd17531@mail.gmail.com> <460D17BA.6070201@netmagicsolutions.com> <57127b5f0703300705o694325f2h5c2b0fac625388eb@mail.gmail.com> Message-ID: <460D4843.7030607@redhat.com> Deependra Shekhawat wrote: > Thanks for the quick response. I am moving forward with my learning. The response below is of course absolutely correct. One thing to point out -- Since you are running DHCP on the cobbler server, this would be an excellent time to try out cobbler's dhcp management. Just edit /var/lib/cobbler/settings to change manage_dhcp to '1', then edit /etc/cobbler/dhcp.template to include the correct dhcp setup information -- just replace any IP addresses that are wrong and leave all the "$" fields as is, as cobbler will template those out for you. Then run "cobbler sync". Cobbler will now be managing your dhcp configurations, so whenever you add a system in cobbler, you can also specify the IP address it will get (optionally) and you won't have to hand edit your dhcpd.conf file again. > > On 3/30/07, *Dhawal Doshy* > wrote: > > Deependra Shekhawat wrote: > > Hello guys, > > > > I have installed EL5 via xen. Now I wanted to try cobbler + koan to > > install EL5. Somewhat like un-attended virtual machine > installation. I > > have installed cobbler and koan from the python code. I did 'cobbler > > check' and it only said about /etc/dhcpd.conf. It says to include > > next-servers line. But why I need it when I have DHCP on the > same server > > as I have xen, cobbler and koan. Other things are setup well. > > But now I don't know how do I proceed. If someone have been > through this > > thing please help. > > From 'man dhcpd.conf' > The next-server statement: > The next-server statement is used to specify the host address of the > server from which the initial boot file (specified in the filename > statement) is to be loaded. Server-name should be a numeric IP address > or a domain name. If no next-server parameter applies to a given > client, > the DHCP server's IP address is used. > > To summarize, you could leave it null if your dhcp/pxe server are the > same, else mention the IP address of the pxe server. If cobbler > insists > on the next-server line, then try using your local IP address. > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > > > > -- > Enjoy Life ! > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From jeevanullas at gmail.com Fri Mar 30 18:01:57 2007 From: jeevanullas at gmail.com (Deependra Shekhawat) Date: Fri, 30 Mar 2007 14:01:57 -0400 Subject: [et-mgmt-tools] Deependra has Tagged you! :) Message-ID: <200703301801.l2UI1vJc012793@mx2.redhat.com> An HTML attachment was scrubbed... URL: From John.Sanabria at ece.uprm.edu Fri Mar 30 18:26:21 2007 From: John.Sanabria at ece.uprm.edu (John Sanabria) Date: Fri, 30 Mar 2007 14:26:21 -0400 (AST) Subject: [et-mgmt-tools] what happen? :( In-Reply-To: <460D4760.8070002@redhat.com> References: <50745.136.145.116.114.1175268447.squirrel@ece.uprm.edu> <460D4760.8070002@redhat.com> Message-ID: <56299.136.145.116.114.1175279181.squirrel@ece.uprm.edu> Hi Michael, i'm sending your response to fedora-xen mailing list for any comments from them. About your answer, the machine where i want to create a new virtual node is the 136.145.116.76. This machine has two virtual nodes up and running, so i think that message is not a hypervisor problem, the hypervisor is working. The other hand, last week i create two additional virtual nodes, in that machine (116.76), the problem appear today. Asides, the distro and profile configuration keep the same. I can't understand why today, provisioning with cobbler stop to work. Thanks a lot to fedora-xen mailing list for your comments and thanks Michael. > John Sanabria wrote: >> Hi, >> >> the last week i had executed the next line command to create a virtual >> nodes in 136.145.116.76 host: >> >> cobbler enchant --virt=yes --address=136.145.116.76 >> --profile="xentemplate-profile" >> >> and it work. >> >> Today, i try again to execute the command and appear that message: >> >> running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias >> nocobblerhostkey' root at 136.145.116.76 wget >> http://136.145.116.114/cobbler/koan-0.2.8-1.noarch.rpm >> Warning: Permanently added 'nocobblerhostkey' (RSA) to the list of known >> hosts. >> --11:18:07-- http://136.145.116.114/cobbler/koan-0.2.8-1.noarch.rpm >> Connecting to 136.145.116.114:80... connected. >> HTTP request sent, awaiting response... 200 OK >> Length: 93410 (91K) [application/x-rpm] >> Saving to: `koan-0.2.8-1.noarch.rpm' >> >> 0K .......... .......... .......... .......... .......... 54% 4.08M >> 0s >> 50K .......... .......... .......... .......... . 100% >> 14.1M=0.01s >> >> 11:18:08 (6.00 MB/s) - `koan-0.2.8-1.noarch.rpm' saved [93410/93410] >> >> returns: 0 >> running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias >> nocobblerhostkey' root at 136.145.116.76 up2date install syslinux >> libvirt-python libvirt >> bash: up2date: command not found >> returns: 127 >> running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias >> nocobblerhostkey' root at 136.145.116.76 yum -y install syslinux >> libvirt-python libvirt >> Loading "installonlyn" plugin >> Setting up Install Process >> Setting up repositories >> Reading repository metadata in from local files >> Parsing package install arguments >> Nothing to do >> returns: 0 >> running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias >> nocobblerhostkey' root at 136.145.116.76 rpm -Uvh koan-0.2.8-1.noarch.rpm >> --force --nodeps >> Preparing... >> ################################################## >> koan >> ################################################## >> returns: 0 >> running: ssh -o 'StrictHostKeyChecking=no' -o 'HostKeyAlias >> nocobblerhostkey' root at 136.145.116.76 koan --virt >> --profile=xentemplate-profile --server=136.145.116.114 >> - processing profile: xentemplate-profile >> - fetching configuration for profile: xentemplate-profile >> - url=http://136.145.116.114/cobbler/profiles/xentemplate-profile >> - {'kickstart': >> 'http://136.145.116.114/cobbler_track/kickstarts/xentemplate-profile/ks.cfg', >> 'name': 'xentemplate-profile', 'virt_ram': 256, 'repos': '', >> 'kernel_options': 'lang= ksdevice=eth0 text syslog=136.145.116.114:25150 >> devfs=nomount ramdisk_size=16438 append ', 'virt_name': 'xentemplate', >> 'virt_file_size': 6, 'distro': 'xentemplate', 'virt_paravirt': 'True', >> 'ks_meta': ''} >> - fetching configuration for distro: xentemplate >> - url=http://136.145.116.114/cobbler/distros/xentemplate >> - {'kernel': '/root/vmlinuz', 'ks_meta': '', 'breed': 'redhat', >> 'kernel_options': 'lang= ksdevice=eth0 text syslog=136.145.116.114:25150 >> devfs=nomount ramdisk_size=16438 append ', 'initrd': '/root/initrd.img', >> 'arch': 'x86', 'name': 'xentemplate'} >> - downloading initrd initrd.img to /tmp/initrd.img >> - url=http://136.145.116.114/cobbler/images/xentemplate/initrd.img >> -libvir: error : no support for hypervisor >> Traceback (most recent call last): >> File "/usr/lib/python2.4/site-packages/koan/app.py", line 97, in main >> k.run() >> File "/usr/lib/python2.4/site-packages/koan/app.py", line 152, in run >> self.do_virt() >> File "/usr/lib/python2.4/site-packages/koan/app.py", line 297, in >> do_virt >> return self.do_net_install("/tmp",after_download) >> File "/usr/lib/python2.4/site-packages/koan/app.py", line 253, in >> do_net_install >> after_download(self, distro_data, profile_data) >> File "/usr/lib/python2.4/site-packages/koan/app.py", line 296, in >> after_download >> self.do_virt_net_install(profile_data, distro_data) >> File "/usr/lib/python2.4/site-packages/koan/app.py", line 600, in >> do_virt_net_install >> extra=kextra >> File "/usr/lib/python2.4/site-packages/koan/virtcreate.py", line 84, >> in >> start_paravirt_install >> guest = virtinst.ParaVirtGuest() >> File "/usr/lib/python2.4/site-packages/virtinst/ParaVirtGuest.py", >> line >> 37, in __init__ >> XenGuest.XenGuest.__init__(self, hypervisorURI=hypervisorURI) >> File "/usr/lib/python2.4/site-packages/virtinst/XenGuest.py", line >> 183, >> in __init__ >> self.conn = libvirt.open(hypervisorURI) >> downloading kernel vmlinuz to /tmp/vmlinuz >> - url=http://136.145.116.114/cobbler/images/xentemplate/vmlinuz >> - kernel saved = /tmp/vmlinuz >> - initrd saved = /tmp/initrd.img >> File "/usr/lib/python2.4/site-packages/libvirt.py", line 95, in open >> if ret is None:raise libvirtError('virConnectOpen() failed') >> libvirtError: virConnectOpen() failed no support for hypervisor >> returns: 3 >> Traceback (most recent call last): >> File "/usr/lib/python2.4/site-packages/cobbler/action_enchant.py", >> line >> 101, in run >> self.ssh_exec("koan %s --profile=%s --server=%s" % (operation, >> self.profile, self.settings.server)) >> File "/usr/lib/python2.4/site-packages/cobbler/action_enchant.py", >> line >> 57, in ssh_exec >> raise cexceptions.CobblerException("enchant_failed","ssh failure") >> CobblerException: 'enchant failed (ssh failure)' >> enchant failed (exception) >> >> ---------------------------- >> >> Any suggestions. >> >> the machine where command were submitted has the next configuration: >> >> uname -a : "Linux reboot 2.6.19-1.2911.6.5.fc6xen #1 SMP Sun Mar 4 >> 16:59:41 EST 2007 i686 i686 i386 GNU/Linux" >> koan: 0.2.7-1.fc6 >> cobbler: 0.4.3-0.fc6 >> >> thanks a lot for your answer. >> >> >> >> >> http://ece.uprm.edu/~s047267 >> http://del.icio.us/josanabr >> http://blog-grid.blogspot.com >> >> _______________________________________________ >> et-mgmt-tools mailing list >> et-mgmt-tools at redhat.com >> https://www.redhat.com/mailman/listinfo/et-mgmt-tools >> > > I think the key is... > > "-libvir: error : no support for hypervisor" > > From the above, it looks like you're getting an error from Xen or > libvirt, rather than from something cobbler/koan related. > > Are you sure you are installing a Xen kernel image vs a plain one? Is > xend running? I'm really not super-up on libvirt error messages, so > perhaps Dan/Hugh can field this. You could always ask fedora-xen as > well. > > --Michael > > > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > http://ece.uprm.edu/~s047267 http://del.icio.us/josanabr http://blog-grid.blogspot.com From prime.provogue at gmail.com Sat Mar 31 01:40:49 2007 From: prime.provogue at gmail.com (niyaz chistee) Date: Sat, 31 Mar 2007 07:10:49 +0530 Subject: [et-mgmt-tools] error while running cobbler import Message-ID: <3293c1d10703301840j3fc46b4atc4e4477056cad24d@mail.gmail.com> Hi, I am trying to configure cobbler to configure my xen virtual machines. I tried to import the installation tree of rhel5 from cobbler using the command: cobbler import --mirror=/rhel5/Dump/ --mirror-name=rhel5 It ran for sometime mirroring my rpm's and in last it ended up with this: sent 2597509927 bytes received 50280 bytes 8644127.14 bytes/sec total size is 2596994302 speedup is 1.00 - creating new distro: rhel5 - creating new profile: rhel5 - running repo update on /var/www/cobbler/ks_mirror/rhel5 - no comps file found: /var/www/cobbler/ks_mirror/rhel5/repodata/comps.xml - modifying existing distro: rhel5 - modifying existing profile: rhel5 - running repo update on /var/www/cobbler/ks_mirror/rhel5 - no comps file found: /var/www/cobbler/ks_mirror/rhel5/repodata/comps.xml - creating new distro: rhel5-xen - creating new profile: rhel5-xen - running repo update on /var/www/cobbler/ks_mirror/rhel5 - no comps file found: /var/www/cobbler/ks_mirror/rhel5/repodata/comps.xml Traceback (most recent call last): File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 502, in main BootCLI(sys.argv).run() File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 102, in run self.curry_args(self.args[1:], self.commands['toplevel']) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 418, in curry_args commands[args[0]](args[1:]) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 312, in import_tree return self.apply_args(args,commands,on_ok) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 406, in apply_args on_ok() File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 311, in on_ok = lambda: go_import() File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 304, in go_import self.temp_mirror_name) File "/usr/lib/python2.4/site-packages/cobbler/api.py", line 152, in import_tree return importer.run() File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", line 82, in run self.guess_kickstarts() File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", line 139, in guess_kickstarts results = self.scan_rpm_filename(rpm) File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", line 186, in scan_rpm_filename major = int(major) ValueError: invalid literal for int(): 5Server I don't know what this is all about. But when I ran cobbler report it showed me this: [root at server cobbler]# cobbler report distro : rhel5 kernel : /var/www/cobbler/ks_mirror/rhel5/images/pxeboot/vmlinuz initrd : /var/www/cobbler/ks_mirror/rhel5/images/pxeboot/initrd.img kernel options : {} architecture : x86 ks metadata : {} breed : redhat distro : rhel5-xen kernel : /var/www/cobbler/ks_mirror/rhel5/images/xen/vmlinuz initrd : /var/www/cobbler/ks_mirror/rhel5/images/xen/initrd.img kernel options : {} architecture : x86 ks metadata : {} breed : redhat profile : rhel5 distro : rhel5 kickstart : /etc/cobbler/default.ks kernel options : {} ks metadata : {} virt file size : 5 virt ram : 512 virt paravirt : True repos : profile : rhel5-xen distro : rhel5-xen kickstart : /etc/cobbler/default.ks kernel options : {} ks metadata : {} virt file size : 5 virt ram : 512 virt paravirt : True repos : That means it has created the distro and profile part. Is that perfect or something is missing. Please help. Regards SysAdmin -------------- next part -------------- An HTML attachment was scrubbed... URL: From prime.provogue at gmail.com Sat Mar 31 02:56:34 2007 From: prime.provogue at gmail.com (niyaz chistee) Date: Sat, 31 Mar 2007 08:26:34 +0530 Subject: [et-mgmt-tools] Re: error while running cobbler import In-Reply-To: <3293c1d10703301840j3fc46b4atc4e4477056cad24d@mail.gmail.com> References: <3293c1d10703301840j3fc46b4atc4e4477056cad24d@mail.gmail.com> Message-ID: <3293c1d10703301956y5fcb1a72ue12fedb5081b04d9@mail.gmail.com> Well I thought that I should go the other way. So I did this: Note: I have my RHEL5 Dump in /rhel5/Dump cobbler distro add --name=rhel5 --kernel=/rhel5/Dump/images/xen/vmlinuz --initrd=/rhel5/Dump/images/initrd.img This ran good. Then I did the following: cobbler profile add --name=redhat5x --distro=rhel5 --kickstart=/rhel5/Dump/ks.cfg --virt-file-size=2 --virt-ram=128 cobbler system add --name=00:16:3E:5F:93:6A --profile=redhat5x Okay Now I exported my /rhel5/Dump via NFS. Here's cat /etc/exports /rhel5/Dump 10.1.1.0/255.255.255.0(rw,sync,no_root_squash) For the sake of convience I am also posting my /rhel5/Dump/ks.cfg and /etc/dhcpd.conf cat /etc/dhcpd.conf # ****************************************************************** # Cobbler managed dhcpd.conf file # generated from cobbler dhcp.conf template (Sat Mar 31 02:31:13 2007) # ****************************************************************** ddns-update-style interim; allow booting; allow bootp; ignore client-updates; set vendorclass = option vendor-class-identifier; subnet 192.168.1.0 netmask 255.255.255.0 { option routers 192.168.1.5; option subnet-mask 255.255.255.0; range dynamic-bootp 192.168.1.100 192.168.1.254; filename "/pxelinux.0"; default-lease-time 21600; max-lease-time 43200; next-server 10.1.1.1; } host label1 { hardware ethernet 00:16:3E:5F:93:6A; next-server 10.1.1.1; } cat /rhel5/Dump/ks.cfg install url --url nfs:10.1.1.1:/rhel5/Dump key lang en_US.UTF-8 network --device eth0 --bootproto dhcp rootpw --iscrypted $1$VwD9nalr$06K0bUawzanX72gNk0es91 firewall --disabled authconfig --enableshadow --enablemd5 selinux --disabled timezone --utc Asia/Calcutta bootloader --location=mbr --driveorder=xvda --append="console=xvc0" # The following is the partition information you requested # Note that any partitions you deleted are not expressed # here so unless you clear all partitions first, this is # not guaranteed to work clearpart --all --drives=xvda part /boot --fstype ext3 --size=100 --ondisk=xvda part pv.2 --size=0 --grow --ondisk=xvda volgroup VolGroup00 --pesize=32768 pv.2 logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 --grow logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=144 --grow --maxsize=288 %packages @admin-tools @base @core @dialup @java @system-tools @text-internet Now I started my portmap,tftp,nfs,httpd services and finally I started koan. koan --virt --server=192.168.1.5 --profile=redhat5x It gave me this: - processing profile: redhat5x - fetching configuration for profile: redhat5x - url=http://192.168.1.5/cobbler/profiles/redhat5x - {'kickstart': 'nfs://10.1.1.1/rhel5/Dump/ks.cfg', 'name': 'redhat5x', 'virt_ram': 128, 'repos': '', 'kernel_options': 'ksdevice=eth0 lang= syslog=10.1.1.1:25150 text ', 'virt_file_size': 2, 'distro': 'rhel5-xen', 'virt_paravirt': 'True', 'ks_meta': ''} - fetching configuration for distro: rhel5-xen - url=http://192.168.1.5/cobbler/distros/rhel5-xen - {'kernel': '/rhel5/Dump/images/xen/vmlinuz', 'ks_meta': '', 'breed': 'redhat', 'kernel_options': 'ksdevice=eth0 lang= syslog=10.1.1.1:25150 text ', 'initrd': '/rhel5/Dump/images/xen/initrd.img', 'arch': 'x86', 'name': 'rhel5-xen'} - downloading initrd initrd.img to /tmp/initrd.img - url=http://192.168.1.5/cobbler/images/rhel5-xen/initrd.img - downloading kernel vmlinuz to /tmp/vmlinuz - url=http://192.168.1.5/cobbler/images/rhel5-xen/vmlinuz - kernel saved = /tmp/vmlinuz - initrd saved = /tmp/initrd.img - invalid RAM size specified, defaulting to 256 MB libvir: Xen Daemon error : POST operation failed: (xend.err 'Error creating domain: The privileged domain did not balloon!') Failed to create domain 00_16_3E_5C_D5_0C Traceback (most recent call last): File "/usr/lib/python2.4/site-packages/koan/app.py", line 97, in main k.run() File "/usr/lib/python2.4/site-packages/koan/app.py", line 152, in run self.do_virt() File "/usr/lib/python2.4/site-packages/koan/app.py", line 297, in do_virt return self.do_net_install("/tmp",after_download) File "/usr/lib/python2.4/site-packages/koan/app.py", line 253, in do_net_install after_download(self, distro_data, profile_data) File "/usr/lib/python2.4/site-packages/koan/app.py", line 296, in after_download self.do_virt_net_install(profile_data, distro_data) File "/usr/lib/python2.4/site-packages/koan/app.py", line 600, in do_virt_net_install extra=kextra File "/usr/lib/python2.4/site-packages/koan/virtcreate.py", line 102, in start_paravirt_install guest.start_install() File "/usr/lib/python2.4/site-packages/virtinst/ParaVirtGuest.py", line 220, in start_install return XenGuest.XenGuest.start_install(self, consolecb) File "/usr/lib/python2.4/site-packages/virtinst/XenGuest.py", line 367, in start_install self.domain = self.conn.createLinux(cxml, 0) File "/usr/lib/python2.4/site-packages/libvirt.py", line 249, in createLinux if ret is None:raise libvirtError('virDomainCreateLinux() failed') libvirtError: virDomainCreateLinux() failed I don't know what happened during the ballon but it didn't created the virtual machine I was expecting. One more thing I would like to through light on is that I have configured a virtual dummy0 network interface which is bridged to xenbr1. Here's my ifconfig(not shown are the xen virtual interfaces): dummy0 Link encap:Ethernet HWaddr F6:42:E2:23:B6:57 inet addr:10.1.1.1 Bcast:10.1.1.255 Mask:255.255.255.0 inet6 addr: fe80::f442:e2ff:fe23:b657/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:41 errors:0 dropped:0 overruns:0 frame:0 TX packets:154 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:7438 (7.2 KiB) TX bytes:11815 (11.5 KiB) eth1 Link encap:Ethernet HWaddr 00:13:02:4B:70:05 inet addr:192.168.1.5 Bcast:192.168.1.255 Mask:255.255.255.0 inet6 addr: fe80::213:2ff:fe4b:7005/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:318 errors:0 dropped:45 overruns:0 frame:0 TX packets:336 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:1000 RX bytes:173630 (169.5 KiB) TX bytes:30388 (29.6 KiB) Interrupt:16 Base address:0xe000 Memory:b0200000-b0200fff All my xen related network stuff is handled by dummy0 and it is then masqureaded to eth1. I think there is problem with /etc/dhcpd.conf regarding eth1 and dummy0. I am not getting it. Please through some light. Regards On 3/31/07, niyaz chistee wrote: > > Hi, > > I am trying to configure cobbler to configure my xen virtual machines. I > tried to import the installation tree of rhel5 from cobbler using the > command: > > cobbler import --mirror=/rhel5/Dump/ --mirror-name=rhel5 > > It ran for sometime mirroring my rpm's and in last it ended up with this: > sent 2597509927 bytes received 50280 bytes 8644127.14 bytes/sec > total size is 2596994302 speedup is 1.00 > - creating new distro: rhel5 > - creating new profile: rhel5 > - running repo update on /var/www/cobbler/ks_mirror/rhel5 > - no comps file found: /var/www/cobbler/ks_mirror/rhel5/repodata/comps.xml > - modifying existing distro: rhel5 > - modifying existing profile: rhel5 > - running repo update on /var/www/cobbler/ks_mirror/rhel5 > - no comps file found: /var/www/cobbler/ks_mirror/rhel5/repodata/comps.xml > - creating new distro: rhel5-xen > - creating new profile: rhel5-xen > - running repo update on /var/www/cobbler/ks_mirror/rhel5 > - no comps file found: /var/www/cobbler/ks_mirror/rhel5/repodata/comps.xml > Traceback (most recent call last): > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 502, in > main > BootCLI( sys.argv).run() > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 102, in > run > self.curry_args(self.args[1:], self.commands['toplevel']) > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 418, in > curry_args > commands[args[0]](args[1:]) > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 312, in > import_tree > return self.apply_args(args,commands,on_ok) > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 406, in > apply_args > on_ok() > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 311, in > > on_ok = lambda: go_import() > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 304, in > go_import > self.temp_mirror_name) > File "/usr/lib/python2.4/site-packages/cobbler/api.py", line 152, in > import_tree > return importer.run() > File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", line > 82, in run > self.guess_kickstarts() > File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", line > 139, in guess_kickstarts > results = self.scan_rpm_filename(rpm) > File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", line > 186, in scan_rpm_filename > major = int(major) > ValueError: invalid literal for int(): 5Server > > I don't know what this is all about. But when I ran cobbler report it > showed me this: > > [root at server cobbler]# cobbler report > distro : rhel5 > kernel : /var/www/cobbler/ks_mirror/rhel5/images/pxeboot/vmlinuz > initrd : > /var/www/cobbler/ks_mirror/rhel5/images/pxeboot/initrd.img > kernel options : {} > architecture : x86 > ks metadata : {} > breed : redhat > > distro : rhel5-xen > kernel : /var/www/cobbler/ks_mirror/rhel5/images/xen/vmlinuz > initrd : /var/www/cobbler/ks_mirror/rhel5/images/xen/initrd.img > kernel options : {} > architecture : x86 > ks metadata : {} > breed : redhat > > profile : rhel5 > distro : rhel5 > kickstart : /etc/cobbler/default.ks > kernel options : {} > ks metadata : {} > virt file size : 5 > virt ram : 512 > virt paravirt : True > repos : > > profile : rhel5-xen > distro : rhel5-xen > kickstart : /etc/cobbler/default.ks > kernel options : {} > ks metadata : {} > virt file size : 5 > virt ram : 512 > virt paravirt : True > repos : > > That means it has created the distro and profile part. Is that perfect or > something is missing. > > Please help. > > Regards > SysAdmin > -------------- next part -------------- An HTML attachment was scrubbed... URL: From prime.provogue at gmail.com Sat Mar 31 04:34:31 2007 From: prime.provogue at gmail.com (niyaz chistee) Date: Sat, 31 Mar 2007 10:04:31 +0530 Subject: [et-mgmt-tools] Re: error while running cobbler import In-Reply-To: <3293c1d10703301956y5fcb1a72ue12fedb5081b04d9@mail.gmail.com> References: <3293c1d10703301840j3fc46b4atc4e4477056cad24d@mail.gmail.com> <3293c1d10703301956y5fcb1a72ue12fedb5081b04d9@mail.gmail.com> Message-ID: <3293c1d10703302134u48851b10iac56a839ffcef128@mail.gmail.com> Hello guys, I somehow overcome the balloon problem. What I did was I delete the system and profile by cobbler system remove --name= cobbler profile remove --name= And I added again the profile but this time I gave --virt-ram-size=256 (as this was where koan reported warning earlier) and then I created system as I did previously. Then I ran koan once again (as I did earlier) but now a new problem. It starts the virtual machine and ask me to connect to it via xm console. I did so but it hangs during acquiring ip address from dhcp. After sometime it ask me to give ip address manually I do so (10.1.1.45/255.255.255.0 gateway 10.1.1.1) but then it reports error in getting nfs:/10.1.1.1/rhel5/Dump/ks.cfg. I have figured out the problem. But don't know how to resolve it. It is the same as I mentioned earlier. I have two network interfaces. One is eth1 (wireless device) and other is dummy0 (virtual network interface). All the communication of xen vm's with the xen dom0 is done via a bridge xenbr1 which is mapped to dummy0. And in the dhcpd.conf there is nothing for the subnet 10.1.1.0 and if I do manually (as of now cobbler is managing my dhcpd.con) it reports error and doesn't start dhcpd service. The error is logged into /var/log/messages and it is shown below: Mar 31 10:05:01 server dhcpd: Internet Systems Consortium DHCP Server V3.0.5-RedHat Mar 31 10:05:01 server dhcpd: Copyright 2004-2006 Internet Systems Consortium. Mar 31 10:05:01 server dhcpd: All rights reserved. Mar 31 10:05:01 server dhcpd: For info, please visit http://www.isc.org/sw/dhcp/ Mar 31 10:05:01 server dhcpd: Wrote 0 deleted host decls to leases file. Mar 31 10:05:01 server dhcpd: Wrote 0 new dynamic host decls to leases file. Mar 31 10:05:01 server dhcpd: Wrote 0 leases to leases file. Mar 31 10:05:01 server dhcpd: Mar 31 10:05:01 server dhcpd: No subnet declaration for eth1 (192.168.1.5). Mar 31 10:05:01 server dhcpd: ** Ignoring requests on eth1. If this is not what Mar 31 10:05:01 server dhcpd: you want, please write a subnet declaration Mar 31 10:05:01 server dhcpd: in your dhcpd.conf file for the network segment Mar 31 10:05:01 server dhcpd: to which interface eth1 is attached. ** Mar 31 10:05:01 server dhcpd: Mar 31 10:05:01 server dhcpd: Mar 31 10:05:01 server dhcpd: Not configured to listen on any interfaces! Mar 31 10:05:01 server dhcpd: Mar 31 10:05:01 server dhcpd: If you did not get this software from ftp.isc.org, please Mar 31 10:05:01 server dhcpd: get the latest from ftp.isc.org and install that before Mar 31 10:05:01 server dhcpd: requesting help. Mar 31 10:05:01 server dhcpd: Mar 31 10:05:01 server dhcpd: If you did get this software from ftp.isc.organd have not Mar 31 10:05:01 server dhcpd: yet read the README, please read it before requesting help. Mar 31 10:05:01 server dhcpd: If you intend to request help from the dhcp-server at isc.org Mar 31 10:05:01 server dhcpd: mailing list, please read the section on the README about Mar 31 10:05:01 server dhcpd: submitting bug reports and requests for help. Mar 31 10:05:01 server dhcpd: Mar 31 10:05:01 server dhcpd: Please do not under any circumstances send requests for Mar 31 10:05:01 server dhcpd: help directly to the authors of this software - please Mar 31 10:05:01 server dhcpd: send them to the appropriate mailing list as described in Mar 31 10:05:01 server dhcpd: the README file. Mar 31 10:05:01 server dhcpd: Mar 31 10:05:01 server dhcpd: exiting. Mar 31 10:05:01 server dhcpd: dhcpd startup failed If you want me to give some more information please feel free to ask. Thanks in advance Regards SysAdmin On 3/31/07, niyaz chistee wrote: > > Well I thought that I should go the other way. So I did this: > Note: I have my RHEL5 Dump in /rhel5/Dump > > cobbler distro add --name=rhel5 --kernel=/rhel5/Dump/images/xen/vmlinuz > --initrd=/rhel5/Dump/images/initrd.img > > This ran good. Then I did the following: > > cobbler profile add --name=redhat5x --distro=rhel5 > --kickstart=/rhel5/Dump/ks.cfg --virt-file-size=2 --virt-ram=128 > cobbler system add --name=00:16:3E:5F:93:6A --profile=redhat5x > > Okay Now I exported my /rhel5/Dump via NFS. Here's cat /etc/exports > /rhel5/Dump 10.1.1.0/255.255.255.0(rw,sync,no_root_squash) > > For the sake of convience I am also posting my /rhel5/Dump/ks.cfg and > /etc/dhcpd.conf > cat /etc/dhcpd.conf > # ****************************************************************** > # Cobbler managed dhcpd.conf file > # generated from cobbler dhcp.conf template (Sat Mar 31 02:31:13 2007) > # ****************************************************************** > ddns-update-style interim; > allow booting; > allow bootp; > ignore client-updates; > set vendorclass = option vendor-class-identifier; > subnet 192.168.1.0 netmask 255.255.255.0 { > option routers 192.168.1.5; > option subnet-mask 255.255.255.0; > range dynamic-bootp 192.168.1.100 192.168.1.254 ; > filename "/pxelinux.0"; > default-lease-time 21600; > max-lease-time 43200; > next-server 10.1.1.1; > } > host label1 { > hardware ethernet 00:16:3E:5F:93:6A; > next-server 10.1.1.1; > } > > cat /rhel5/Dump/ks.cfg > install > url --url nfs:10.1.1.1:/rhel5/Dump > key > lang en_US.UTF-8 > network --device eth0 --bootproto dhcp > rootpw --iscrypted $1$VwD9nalr$06K0bUawzanX72gNk0es91 > firewall --disabled > authconfig --enableshadow --enablemd5 > selinux --disabled > timezone --utc Asia/Calcutta > bootloader --location=mbr --driveorder=xvda --append="console=xvc0" > # The following is the partition information you requested > # Note that any partitions you deleted are not expressed > # here so unless you clear all partitions first, this is > # not guaranteed to work > clearpart --all --drives=xvda > part /boot --fstype ext3 --size=100 --ondisk=xvda > part pv.2 --size=0 --grow --ondisk=xvda > volgroup VolGroup00 --pesize=32768 pv.2 > logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 > --grow > logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=144 > --grow --maxsize=288 > %packages > @admin-tools > @base > @core > @dialup > @java > @system-tools > @text-internet > > Now I started my portmap,tftp,nfs,httpd services and finally I started > koan. > > koan --virt --server=192.168.1.5 --profile=redhat5x > It gave me this: > - processing profile: redhat5x > - fetching configuration for profile: redhat5x > - url= http://192.168.1.5/cobbler/profiles/redhat5x > - {'kickstart': 'nfs://10.1.1.1/rhel5/Dump/ks.cfg', 'name': 'redhat5x', > 'virt_ram': 128, 'repos': '', 'kernel_options': 'ksdevice=eth0 lang= > syslog= 10.1.1.1:25150 text ', 'virt_file_size': 2, 'distro': 'rhel5-xen', > 'virt_paravirt': 'True', 'ks_meta': ''} > - fetching configuration for distro: rhel5-xen > - url=http://192.168.1.5/cobbler/distros/rhel5-xen > - {'kernel': '/rhel5/Dump/images/xen/vmlinuz', 'ks_meta': '', 'breed': > 'redhat', 'kernel_options': 'ksdevice=eth0 lang= syslog= 10.1.1.1:25150text ', 'initrd': '/rhel5/Dump/images/xen/initrd.img', 'arch': 'x86', > 'name': 'rhel5-xen'} > - downloading initrd initrd.img to /tmp/initrd.img > - url=http://192.168.1.5/cobbler/images/rhel5-xen/initrd.img > - downloading kernel vmlinuz to /tmp/vmlinuz > - url= http://192.168.1.5/cobbler/images/rhel5-xen/vmlinuz > - kernel saved = /tmp/vmlinuz > - initrd saved = /tmp/initrd.img > - invalid RAM size specified, defaulting to 256 MB > libvir: Xen Daemon error : POST operation failed: ( xend.err 'Error > creating domain: The privileged domain did not balloon!') > Failed to create domain 00_16_3E_5C_D5_0C > Traceback (most recent call last): > File "/usr/lib/python2.4/site-packages/koan/app.py", line 97, in main > k.run() > File "/usr/lib/python2.4/site-packages/koan/app.py", line 152, in run > self.do_virt() > File "/usr/lib/python2.4/site-packages/koan/app.py", line 297, in > do_virt > return self.do_net_install("/tmp",after_download) > File "/usr/lib/python2.4/site-packages/koan/app.py", line 253, in > do_net_install > after_download(self, distro_data, profile_data) > File "/usr/lib/python2.4/site-packages/koan/app.py", line 296, in > after_download > self.do_virt_net_install(profile_data, distro_data) > File "/usr/lib/python2.4/site-packages/koan/app.py", line 600, in > do_virt_net_install > extra=kextra > File "/usr/lib/python2.4/site-packages/koan/virtcreate.py", line 102, in > start_paravirt_install > guest.start_install() > File "/usr/lib/python2.4/site-packages/virtinst/ParaVirtGuest.py", line > 220, in start_install > return XenGuest.XenGuest.start_install(self, consolecb) > File "/usr/lib/python2.4/site-packages/virtinst/XenGuest.py", line 367, > in start_install > self.domain = self.conn.createLinux(cxml, 0) > File "/usr/lib/python2.4/site-packages/libvirt.py", line 249, in > createLinux > if ret is None:raise libvirtError('virDomainCreateLinux() failed') > libvirtError: virDomainCreateLinux() failed > > I don't know what happened during the ballon but it didn't created the > virtual machine I was expecting. > One more thing I would like to through light on is that I have configured > a virtual dummy0 network interface which is bridged to xenbr1. Here's my > ifconfig(not shown are the xen virtual interfaces): > dummy0 Link encap:Ethernet HWaddr F6:42:E2:23:B6:57 > inet addr:10.1.1.1 Bcast:10.1.1.255 Mask:255.255.255.0 > inet6 addr: fe80::f442:e2ff:fe23:b657/64 Scope:Link > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 > RX packets:41 errors:0 dropped:0 overruns:0 frame:0 > TX packets:154 errors:0 dropped:0 overruns:0 carrier:0 > collisions:0 txqueuelen:0 > RX bytes:7438 (7.2 KiB) TX bytes:11815 (11.5 KiB) > eth1 Link encap:Ethernet HWaddr 00:13:02:4B:70:05 > inet addr:192.168.1.5 Bcast:192.168.1.255 Mask:255.255.255.0 > inet6 addr: fe80::213:2ff:fe4b:7005/64 Scope:Link > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 > RX packets:318 errors:0 dropped:45 overruns:0 frame:0 > TX packets:336 errors:0 dropped:0 overruns:0 carrier:0 > collisions:0 txqueuelen:1000 > RX bytes:173630 (169.5 KiB) TX bytes:30388 ( 29.6 KiB) > Interrupt:16 Base address:0xe000 Memory:b0200000-b0200fff > > All my xen related network stuff is handled by dummy0 and it is then > masqureaded to eth1. I think there is problem with /etc/dhcpd.conf regarding > eth1 and dummy0. I am not getting it. > Please through some light. > > Regards > > On 3/31/07, niyaz chistee wrote: > > > > Hi, > > > > I am trying to configure cobbler to configure my xen virtual machines. I > > tried to import the installation tree of rhel5 from cobbler using the > > command: > > > > cobbler import --mirror=/rhel5/Dump/ --mirror-name=rhel5 > > > > It ran for sometime mirroring my rpm's and in last it ended up with > > this: > > sent 2597509927 bytes received 50280 bytes 8644127.14 bytes/sec > > total size is 2596994302 speedup is 1.00 > > - creating new distro: rhel5 > > - creating new profile: rhel5 > > - running repo update on /var/www/cobbler/ks_mirror/rhel5 > > - no comps file found: > > /var/www/cobbler/ks_mirror/rhel5/repodata/comps.xml > > - modifying existing distro: rhel5 > > - modifying existing profile: rhel5 > > - running repo update on /var/www/cobbler/ks_mirror/rhel5 > > - no comps file found: > > /var/www/cobbler/ks_mirror/rhel5/repodata/comps.xml > > - creating new distro: rhel5-xen > > - creating new profile: rhel5-xen > > - running repo update on /var/www/cobbler/ks_mirror/rhel5 > > - no comps file found: > > /var/www/cobbler/ks_mirror/rhel5/repodata/comps.xml > > Traceback (most recent call last): > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 502, > > in main > > BootCLI( sys.argv).run() > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 102, > > in run > > self.curry_args(self.args[1:], self.commands['toplevel']) > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 418, > > in curry_args > > commands[args[0]](args[1:]) > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 312, > > in import_tree > > return self.apply_args(args,commands,on_ok) > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 406, > > in apply_args > > on_ok() > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 311, > > in > > on_ok = lambda: go_import() > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 304, > > in go_import > > self.temp_mirror_name) > > File "/usr/lib/python2.4/site-packages/cobbler/api.py", line 152, in > > import_tree > > return importer.run() > > File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", line > > 82, in run > > self.guess_kickstarts() > > File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", line > > 139, in guess_kickstarts > > results = self.scan_rpm_filename(rpm) > > File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", line > > 186, in scan_rpm_filename > > major = int(major) > > ValueError: invalid literal for int(): 5Server > > > > I don't know what this is all about. But when I ran cobbler report it > > showed me this: > > > > [root at server cobbler]# cobbler report > > distro : rhel5 > > kernel : > > /var/www/cobbler/ks_mirror/rhel5/images/pxeboot/vmlinuz > > initrd : > > /var/www/cobbler/ks_mirror/rhel5/images/pxeboot/initrd.img > > kernel options : {} > > architecture : x86 > > ks metadata : {} > > breed : redhat > > > > distro : rhel5-xen > > kernel : /var/www/cobbler/ks_mirror/rhel5/images/xen/vmlinuz > > initrd : /var/www/cobbler/ks_mirror/rhel5/images/xen/initrd.img > > > > kernel options : {} > > architecture : x86 > > ks metadata : {} > > breed : redhat > > > > profile : rhel5 > > distro : rhel5 > > kickstart : /etc/cobbler/default.ks > > kernel options : {} > > ks metadata : {} > > virt file size : 5 > > virt ram : 512 > > virt paravirt : True > > repos : > > > > profile : rhel5-xen > > distro : rhel5-xen > > kickstart : /etc/cobbler/default.ks > > kernel options : {} > > ks metadata : {} > > virt file size : 5 > > virt ram : 512 > > virt paravirt : True > > repos : > > > > That means it has created the distro and profile part. Is that perfect > > or something is missing. > > > > Please help. > > > > Regards > > SysAdmin > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From prime.provogue at gmail.com Sat Mar 31 05:59:27 2007 From: prime.provogue at gmail.com (niyaz chistee) Date: Sat, 31 Mar 2007 11:29:27 +0530 Subject: [et-mgmt-tools] Re: error while running cobbler import In-Reply-To: <3293c1d10703302134u48851b10iac56a839ffcef128@mail.gmail.com> References: <3293c1d10703301840j3fc46b4atc4e4477056cad24d@mail.gmail.com> <3293c1d10703301956y5fcb1a72ue12fedb5081b04d9@mail.gmail.com> <3293c1d10703302134u48851b10iac56a839ffcef128@mail.gmail.com> Message-ID: <3293c1d10703302259x12120c73rc789e452c324c18d@mail.gmail.com> Hello, Finally I got rid of the dhcp problem. But I had to stop my dummy0 virtual network interface in dom0 host and have to bridge my xenbr1 to eth1 (physical wireless device on dom0 host). The installation started without asking me anything (that was real great un-attended installation). The only thing it asked was when it was about to create the partitions. It asked me whether it should format the whole disk or not. When I pressed Yes it continued. If I want it to not ask this thing what should I do. And one more thing after installation it stop and asked me to press reboot button. If I want it to automatically reboot then what should I do. I guess something %post directive. Please help. And one more thing, I had to made huge modifications in my kickstart file. So I am posting it once more. Hope it help you guys in getting me: install nfs --server=192.168.1.5 --dir=/rhel5/Dump key lang en_US.UTF-8 network --bootproto=bootp --device=eth0 --onboot=on rootpw --iscrypted $1$VwD9nalr$06K0bUawzanX72gNk0es91 firewall --disabled authconfig --enableshadow --enablemd5 selinux --disabled timezone --utc Asia/Calcutta bootloader --location=mbr --driveorder=xvda --append="console=xvc0" # The following is the partition information you requested # Note that any partitions you deleted are not expressed # here so unless you clear all partitions first, this is # not guaranteed to work clearpart --all --drives=xvda part /boot --fstype ext3 --size=100 --ondisk=xvda part pv.2 --size=0 --grow --ondisk=xvda volgroup VolGroup00 --pesize=32768 pv.2 logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 --grow logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=144 --grow --maxsize=288 %packages @admin-tools @base @core On 3/31/07, niyaz chistee wrote: > > Hello guys, > > I somehow overcome the balloon problem. What I did was I delete the system > and profile by > cobbler system remove --name= > cobbler profile remove --name= > > And I added again the profile but this time I gave --virt-ram-size=256 (as > this was where koan reported warning earlier) and then I created system as > I did previously. > Then I ran koan once again (as I did earlier) but now a new problem. It > starts the virtual machine and ask me to connect to it via xm console. I did > so but it hangs during acquiring ip address from dhcp. After sometime it ask > me to give ip address manually I do so ( 10.1.1.45/255.255.255.0 gateway > 10.1.1.1) but then it reports error in getting > nfs:/10.1.1.1/rhel5/Dump/ks.cfg. I have figured out the problem. But don't > know how to resolve it. > It is the same as I mentioned earlier. I have two network interfaces. One > is eth1 (wireless device) and other is dummy0 (virtual network interface). > All the communication of xen vm's with the xen dom0 is done via a bridge > xenbr1 which is mapped to dummy0. And in the dhcpd.conf there is nothing > for the subnet 10.1.1.0 and if I do manually (as of now cobbler is > managing my dhcpd.con) it reports error and doesn't start dhcpd service. > The error is logged into /var/log/messages and it is shown below: > > Mar 31 10:05:01 server dhcpd: Internet Systems Consortium DHCP Server > V3.0.5-RedHat > Mar 31 10:05:01 server dhcpd: Copyright 2004-2006 Internet Systems > Consortium. > Mar 31 10:05:01 server dhcpd: All rights reserved. > Mar 31 10:05:01 server dhcpd: For info, please visit > http://www.isc.org/sw/dhcp/ > Mar 31 10:05:01 server dhcpd: Wrote 0 deleted host decls to leases file. > Mar 31 10:05:01 server dhcpd: Wrote 0 new dynamic host decls to leases > file. > Mar 31 10:05:01 server dhcpd: Wrote 0 leases to leases file. > Mar 31 10:05:01 server dhcpd: > Mar 31 10:05:01 server dhcpd: No subnet declaration for eth1 (192.168.1.5 > ). > Mar 31 10:05:01 server dhcpd: ** Ignoring requests on eth1. If this is > not what > Mar 31 10:05:01 server dhcpd: you want, please write a subnet > declaration > Mar 31 10:05:01 server dhcpd: in your dhcpd.conf file for the network > segment > Mar 31 10:05:01 server dhcpd: to which interface eth1 is attached. ** > Mar 31 10:05:01 server dhcpd: > Mar 31 10:05:01 server dhcpd: > Mar 31 10:05:01 server dhcpd: Not configured to listen on any interfaces! > Mar 31 10:05:01 server dhcpd: > Mar 31 10:05:01 server dhcpd: If you did not get this software from > ftp.isc.org, please > Mar 31 10:05:01 server dhcpd: get the latest from ftp.isc.org and install > that before > Mar 31 10:05:01 server dhcpd: requesting help. > Mar 31 10:05:01 server dhcpd: > Mar 31 10:05:01 server dhcpd: If you did get this software from > ftp.isc.org and have not > Mar 31 10:05:01 server dhcpd: yet read the README, please read it before > requesting help. > Mar 31 10:05:01 server dhcpd: If you intend to request help from the > dhcp-server at isc.org > Mar 31 10:05:01 server dhcpd: mailing list, please read the section on the > README about > Mar 31 10:05:01 server dhcpd: submitting bug reports and requests for > help. > Mar 31 10:05:01 server dhcpd: > Mar 31 10:05:01 server dhcpd: Please do not under any circumstances send > requests for > Mar 31 10:05:01 server dhcpd: help directly to the authors of this > software - please > Mar 31 10:05:01 server dhcpd: send them to the appropriate mailing list as > described in > Mar 31 10:05:01 server dhcpd: the README file. > Mar 31 10:05:01 server dhcpd: > Mar 31 10:05:01 server dhcpd: exiting. > Mar 31 10:05:01 server dhcpd: dhcpd startup failed > > If you want me to give some more information please feel free to ask. > Thanks in advance > > Regards > SysAdmin > > On 3/31/07, niyaz chistee wrote: > > > > Well I thought that I should go the other way. So I did this: > > Note: I have my RHEL5 Dump in /rhel5/Dump > > > > cobbler distro add --name=rhel5 --kernel=/rhel5/Dump/images/xen/vmlinuz > > --initrd=/rhel5/Dump/images/initrd.img > > > > This ran good. Then I did the following: > > > > cobbler profile add --name=redhat5x --distro=rhel5 > > --kickstart=/rhel5/Dump/ks.cfg --virt-file-size=2 --virt-ram=128 > > cobbler system add --name=00:16:3E:5F:93:6A --profile=redhat5x > > > > Okay Now I exported my /rhel5/Dump via NFS. Here's cat /etc/exports > > /rhel5/Dump 10.1.1.0/255.255.255.0(rw,sync,no_root_squash) > > > > For the sake of convience I am also posting my /rhel5/Dump/ks.cfg and > > /etc/dhcpd.conf > > cat /etc/dhcpd.conf > > # ****************************************************************** > > # Cobbler managed dhcpd.conf file > > # generated from cobbler dhcp.conf template (Sat Mar 31 02:31:13 2007) > > # ****************************************************************** > > ddns-update-style interim; > > allow booting; > > allow bootp; > > ignore client-updates; > > set vendorclass = option vendor-class-identifier; > > subnet 192.168.1.0 netmask 255.255.255.0 { > > option routers 192.168.1.5; > > option subnet-mask 255.255.255.0; > > range dynamic-bootp 192.168.1.100 192.168.1.254 ; > > filename "/pxelinux.0"; > > default-lease-time 21600; > > max-lease-time 43200; > > next-server 10.1.1.1; > > } > > host label1 { > > hardware ethernet 00:16:3E:5F:93:6A; > > next-server 10.1.1.1; > > } > > > > cat /rhel5/Dump/ks.cfg > > install > > url --url nfs:10.1.1.1:/rhel5/Dump > > key > > lang en_US.UTF-8 > > network --device eth0 --bootproto dhcp > > rootpw --iscrypted $1$VwD9nalr$06K0bUawzanX72gNk0es91 > > firewall --disabled > > authconfig --enableshadow --enablemd5 > > selinux --disabled > > timezone --utc Asia/Calcutta > > bootloader --location=mbr --driveorder=xvda --append="console=xvc0" > > # The following is the partition information you requested > > # Note that any partitions you deleted are not expressed > > # here so unless you clear all partitions first, this is > > # not guaranteed to work > > clearpart --all --drives=xvda > > part /boot --fstype ext3 --size=100 --ondisk=xvda > > part pv.2 --size=0 --grow --ondisk=xvda > > volgroup VolGroup00 --pesize=32768 pv.2 > > logvol / --fstype ext3 --name=LogVol00 --vgname=VolGroup00 --size=1024 > > --grow > > logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=144 > > --grow --maxsize=288 > > %packages > > @admin-tools > > @base > > @core > > @dialup > > @java > > @system-tools > > @text-internet > > > > Now I started my portmap,tftp,nfs,httpd services and finally I started > > koan. > > > > koan --virt --server=192.168.1.5 --profile=redhat5x > > It gave me this: > > - processing profile: redhat5x > > - fetching configuration for profile: redhat5x > > - url= http://192.168.1.5/cobbler/profiles/redhat5x > > - {'kickstart': 'nfs://10.1.1.1/rhel5/Dump/ks.cfg', 'name': 'redhat5x', > > 'virt_ram': 128, 'repos': '', 'kernel_options': 'ksdevice=eth0 lang= > > syslog= 10.1.1.1:25150 text ', 'virt_file_size': 2, 'distro': > > 'rhel5-xen', 'virt_paravirt': 'True', 'ks_meta': ''} > > - fetching configuration for distro: rhel5-xen > > - url=http://192.168.1.5/cobbler/distros/rhel5-xen > > - {'kernel': '/rhel5/Dump/images/xen/vmlinuz', 'ks_meta': '', 'breed': > > 'redhat', 'kernel_options': 'ksdevice=eth0 lang= syslog= 10.1.1.1:25150text ', 'initrd': '/rhel5/Dump/images/xen/initrd.img', 'arch': 'x86', > > 'name': 'rhel5-xen'} > > - downloading initrd initrd.img to /tmp/initrd.img > > - url=http://192.168.1.5/cobbler/images/rhel5-xen/initrd.img > > - downloading kernel vmlinuz to /tmp/vmlinuz > > - url= http://192.168.1.5/cobbler/images/rhel5-xen/vmlinuz > > - kernel saved = /tmp/vmlinuz > > - initrd saved = /tmp/initrd.img > > - invalid RAM size specified, defaulting to 256 MB > > libvir: Xen Daemon error : POST operation failed: ( xend.err 'Error > > creating domain: The privileged domain did not balloon!') > > Failed to create domain 00_16_3E_5C_D5_0C > > Traceback (most recent call last): > > File "/usr/lib/python2.4/site-packages/koan/app.py", line 97, in main > > k.run() > > File "/usr/lib/python2.4/site-packages/koan/app.py", line 152, in run > > self.do_virt() > > File "/usr/lib/python2.4/site-packages/koan/app.py", line 297, in > > do_virt > > return self.do_net_install("/tmp",after_download) > > File "/usr/lib/python2.4/site-packages/koan/app.py", line 253, in > > do_net_install > > after_download(self, distro_data, profile_data) > > File "/usr/lib/python2.4/site-packages/koan/app.py", line 296, in > > after_download > > self.do_virt_net_install(profile_data, distro_data) > > File "/usr/lib/python2.4/site-packages/koan/app.py", line 600, in > > do_virt_net_install > > extra=kextra > > File "/usr/lib/python2.4/site-packages/koan/virtcreate.py", line 102, > > in start_paravirt_install > > guest.start_install() > > File "/usr/lib/python2.4/site-packages/virtinst/ParaVirtGuest.py", > > line 220, in start_install > > return XenGuest.XenGuest.start_install(self, consolecb) > > File "/usr/lib/python2.4/site-packages/virtinst/XenGuest.py", line > > 367, in start_install > > self.domain = self.conn.createLinux(cxml, 0) > > File "/usr/lib/python2.4/site-packages/libvirt.py", line 249, in > > createLinux > > if ret is None:raise libvirtError('virDomainCreateLinux() failed') > > libvirtError: virDomainCreateLinux() failed > > > > I don't know what happened during the ballon but it didn't created the > > virtual machine I was expecting. > > One more thing I would like to through light on is that I have > > configured a virtual dummy0 network interface which is bridged to xenbr1. > > Here's my ifconfig(not shown are the xen virtual interfaces): > > dummy0 Link encap:Ethernet HWaddr F6:42:E2:23:B6:57 > > inet addr:10.1.1.1 Bcast: 10.1.1.255 Mask:255.255.255.0 > > inet6 addr: fe80::f442:e2ff:fe23:b657/64 Scope:Link > > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 > > RX packets:41 errors:0 dropped:0 overruns:0 frame:0 > > TX packets:154 errors:0 dropped:0 overruns:0 carrier:0 > > collisions:0 txqueuelen:0 > > RX bytes:7438 (7.2 KiB) TX bytes:11815 (11.5 KiB) > > eth1 Link encap:Ethernet HWaddr 00:13:02:4B:70:05 > > inet addr: 192.168.1.5 Bcast:192.168.1.255 Mask:255.255.255.0 > > inet6 addr: fe80::213:2ff:fe4b:7005/64 Scope:Link > > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 > > RX packets:318 errors:0 dropped:45 overruns:0 frame:0 > > TX packets:336 errors:0 dropped:0 overruns:0 carrier:0 > > collisions:0 txqueuelen:1000 > > RX bytes:173630 (169.5 KiB) TX bytes:30388 ( 29.6 KiB) > > Interrupt:16 Base address:0xe000 Memory:b0200000-b0200fff > > > > All my xen related network stuff is handled by dummy0 and it is then > > masqureaded to eth1. I think there is problem with /etc/dhcpd.conf regarding > > eth1 and dummy0. I am not getting it. > > Please through some light. > > > > Regards > > > > On 3/31/07, niyaz chistee < prime.provogue at gmail.com> wrote: > > > > > > Hi, > > > > > > I am trying to configure cobbler to configure my xen virtual machines. > > > I tried to import the installation tree of rhel5 from cobbler using the > > > command: > > > > > > cobbler import --mirror=/rhel5/Dump/ --mirror-name=rhel5 > > > > > > It ran for sometime mirroring my rpm's and in last it ended up with > > > this: > > > sent 2597509927 bytes received 50280 bytes 8644127.14 bytes/sec > > > total size is 2596994302 speedup is 1.00 > > > - creating new distro: rhel5 > > > - creating new profile: rhel5 > > > - running repo update on /var/www/cobbler/ks_mirror/rhel5 > > > - no comps file found: > > > /var/www/cobbler/ks_mirror/rhel5/repodata/comps.xml > > > - modifying existing distro: rhel5 > > > - modifying existing profile: rhel5 > > > - running repo update on /var/www/cobbler/ks_mirror/rhel5 > > > - no comps file found: > > > /var/www/cobbler/ks_mirror/rhel5/repodata/comps.xml > > > - creating new distro: rhel5-xen > > > - creating new profile: rhel5-xen > > > - running repo update on /var/www/cobbler/ks_mirror/rhel5 > > > - no comps file found: > > > /var/www/cobbler/ks_mirror/rhel5/repodata/comps.xml > > > Traceback (most recent call last): > > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line > > > 502, in main > > > BootCLI( sys.argv).run() > > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line > > > 102, in run > > > self.curry_args(self.args[1:], self.commands['toplevel']) > > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line > > > 418, in curry_args > > > commands[args[0]](args[1:]) > > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line > > > 312, in import_tree > > > return self.apply_args(args,commands,on_ok) > > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line > > > 406, in apply_args > > > on_ok() > > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line > > > 311, in > > > on_ok = lambda: go_import() > > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line > > > 304, in go_import > > > self.temp_mirror_name) > > > File "/usr/lib/python2.4/site-packages/cobbler/api.py", line 152, in > > > import_tree > > > return importer.run() > > > File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", > > > line 82, in run > > > self.guess_kickstarts() > > > File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", > > > line 139, in guess_kickstarts > > > results = self.scan_rpm_filename(rpm) > > > File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", > > > line 186, in scan_rpm_filename > > > major = int(major) > > > ValueError: invalid literal for int(): 5Server > > > > > > I don't know what this is all about. But when I ran cobbler report it > > > showed me this: > > > > > > [root at server cobbler]# cobbler report > > > distro : rhel5 > > > kernel : > > > /var/www/cobbler/ks_mirror/rhel5/images/pxeboot/vmlinuz > > > initrd : > > > /var/www/cobbler/ks_mirror/rhel5/images/pxeboot/initrd.img > > > kernel options : {} > > > architecture : x86 > > > ks metadata : {} > > > breed : redhat > > > > > > distro : rhel5-xen > > > kernel : /var/www/cobbler/ks_mirror/rhel5/images/xen/vmlinuz > > > initrd : > > > /var/www/cobbler/ks_mirror/rhel5/images/xen/initrd.img > > > kernel options : {} > > > architecture : x86 > > > ks metadata : {} > > > breed : redhat > > > > > > profile : rhel5 > > > distro : rhel5 > > > kickstart : /etc/cobbler/default.ks > > > kernel options : {} > > > ks metadata : {} > > > virt file size : 5 > > > virt ram : 512 > > > virt paravirt : True > > > repos : > > > > > > profile : rhel5-xen > > > distro : rhel5-xen > > > kickstart : /etc/cobbler/default.ks > > > kernel options : {} > > > ks metadata : {} > > > virt file size : 5 > > > virt ram : 512 > > > virt paravirt : True > > > repos : > > > > > > That means it has created the distro and profile part. Is that perfect > > > or something is missing. > > > > > > Please help. > > > > > > Regards > > > SysAdmin > > > > > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From prime.provogue at gmail.com Sat Mar 31 10:15:55 2007 From: prime.provogue at gmail.com (niyaz chistee) Date: Sat, 31 Mar 2007 15:45:55 +0530 Subject: [et-mgmt-tools] Kernel panic while booting rhel4 via pxe Message-ID: <3293c1d10703310315n2669ff34lbe681801554ab116@mail.gmail.com> Hello guys, Once again. After successfully installing a virtual xen machine via cobbler + koan I moved a step forward in installing a new physical machine via cobbler + koan. I have installed cobbler and koan on EL4. All is set well. Now when I boot the client it gets ip address from the server and show my profile (rhel4x) in the pxe boot menu. Now when I write this: boot: rhel4x and boot the system after doing some initialization it fails and shows this: VFS: Cannot open root device "" or unknown-block(104,3) Please append a correct "root=" boot option Kernel panic - not syncing: VFS: Unable to mount root fs on unknown-block(104,3) Well. I was really not expecting this as everything was fine. But it came. What should I do? Please list members do reply to my email this time. Regards SysAdmin -------------- next part -------------- An HTML attachment was scrubbed... URL: From dhawal at netmagicsolutions.com Sat Mar 31 10:39:39 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Sat, 31 Mar 2007 16:09:39 +0530 Subject: [et-mgmt-tools] Kernel panic while booting rhel4 via pxe In-Reply-To: <3293c1d10703310315n2669ff34lbe681801554ab116@mail.gmail.com> References: <3293c1d10703310315n2669ff34lbe681801554ab116@mail.gmail.com> Message-ID: <460E3A6B.3090702@netmagicsolutions.com> niyaz chistee wrote: > Hello guys, > > Once again. After successfully installing a virtual xen machine via > cobbler + koan I moved a step forward in installing a new physical > machine via cobbler + koan. > I have installed cobbler and koan on EL4. All is set well. Now when I > boot the client it gets ip address from the server and show my profile > (rhel4x) in the pxe boot menu. Now when I write this: > boot: rhel4x > and boot the system after doing some initialization it fails and shows this: > > VFS: Cannot open root device "" or unknown-block(104,3) > Please append a correct "root=" boot option > Kernel panic - not syncing: VFS: Unable to mount root fs on > unknown-block(104,3) > > Well. I was really not expecting this as everything was fine. But it > came. What should I do? > Please list members do reply to my email this time. it would help if you post your relevant 'cobbler list/report' and content of 'pxelinux.cfg/default' From prime.provogue at gmail.com Sat Mar 31 10:51:39 2007 From: prime.provogue at gmail.com (niyaz chistee) Date: Sat, 31 Mar 2007 16:21:39 +0530 Subject: [et-mgmt-tools] Kernel panic while booting rhel4 via pxe In-Reply-To: <460E3A6B.3090702@netmagicsolutions.com> References: <3293c1d10703310315n2669ff34lbe681801554ab116@mail.gmail.com> <460E3A6B.3090702@netmagicsolutions.com> Message-ID: <3293c1d10703310351i3d0b45dbxa42124be3ddee136@mail.gmail.com> Thanks for quick response. This is my cobbler --report: distro : rhel4 kernel : /media/hda6/rhel4/images/pxeboot/vmlinuz initrd : /media/hda6/rhel4/images/pxeboot/initrd.img kernel options : {} architecture : x86 ks metadata : {} breed : redhat profile : rhel4x distro : rhel4 kickstart : /media/hda6/rhel4/ks.cfg kernel options : {} ks metadata : {} virt file size : 5 virt ram : 512 virt paravirt : True repos : And here is my /tftpboot/pxelinux.cfg/default: DEFAULT local PROMPT 1 MENU TITLE Cobbler | http://cobbler.et.redhat.com TIMEOUT 200 TOTALTIMEOUT 6000 ONTIMEOUT local LABEL local MENU LABEL (local) MENU DEFAULT LOCALBOOT 0 LABEL rhel4x kernel /images/rhel4/vmlinuz MENU LABEL rhel4x append ksdevice=eth0 lang= syslog=192.168.1.45:25150 text initrd=/images/rhel4/initrd.img ks= http://192.168.1.45/cblr/kickstarts/rhel4x/ks.cfg Hope it helps. Thanks in advance. Regards SysAdmin On 3/31/07, Dhawal Doshy wrote: > > niyaz chistee wrote: > > Hello guys, > > > > Once again. After successfully installing a virtual xen machine via > > cobbler + koan I moved a step forward in installing a new physical > > machine via cobbler + koan. > > I have installed cobbler and koan on EL4. All is set well. Now when I > > boot the client it gets ip address from the server and show my profile > > (rhel4x) in the pxe boot menu. Now when I write this: > > boot: rhel4x > > and boot the system after doing some initialization it fails and shows > this: > > > > VFS: Cannot open root device "" or unknown-block(104,3) > > Please append a correct "root=" boot option > > Kernel panic - not syncing: VFS: Unable to mount root fs on > > unknown-block(104,3) > > > > Well. I was really not expecting this as everything was fine. But it > > came. What should I do? > > Please list members do reply to my email this time. > > it would help if you post your relevant 'cobbler list/report' and > content of 'pxelinux.cfg/default' > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > -------------- next part -------------- An HTML attachment was scrubbed... URL: From dhawal at netmagicsolutions.com Sat Mar 31 11:17:09 2007 From: dhawal at netmagicsolutions.com (Dhawal Doshy) Date: Sat, 31 Mar 2007 16:47:09 +0530 Subject: [et-mgmt-tools] Kernel panic while booting rhel4 via pxe In-Reply-To: <3293c1d10703310351i3d0b45dbxa42124be3ddee136@mail.gmail.com> References: <3293c1d10703310315n2669ff34lbe681801554ab116@mail.gmail.com> <460E3A6B.3090702@netmagicsolutions.com> <3293c1d10703310351i3d0b45dbxa42124be3ddee136@mail.gmail.com> Message-ID: <460E4335.5060503@netmagicsolutions.com> niyaz chistee wrote: > Thanks for quick response. This is my cobbler --report: > distro : rhel4 > kernel : /media/hda6/rhel4/images/pxeboot/vmlinuz > initrd : /media/hda6/rhel4/images/pxeboot/initrd.img > kernel options : {} > architecture : x86 > ks metadata : {} > breed : redhat > > profile : rhel4x > distro : rhel4 > kickstart : /media/hda6/rhel4/ks.cfg > kernel options : {} > ks metadata : {} > virt file size : 5 > virt ram : 512 > virt paravirt : True > repos : > > And here is my /tftpboot/pxelinux.cfg/default: > > DEFAULT local > PROMPT 1 > MENU TITLE Cobbler | http://cobbler.et.redhat.com > TIMEOUT 200 > TOTALTIMEOUT 6000 > ONTIMEOUT local > > LABEL local > MENU LABEL (local) > MENU DEFAULT > LOCALBOOT 0 > > LABEL rhel4x > kernel /images/rhel4/vmlinuz > MENU LABEL rhel4x > append ksdevice=eth0 lang= syslog=192.168.1.45:25150 > text initrd=/images/rhel4/initrd.img ks= > http://192.168.1.45/cblr/kickstarts/rhel4x/ks.cfg You do not appear to have a ramdisk_size=8192 OR ramdisk_size=16438 in your 'append' line of '/tftpboot/pxelinux.cfg/default'. Maybe (just maybe) that is the cause of the this problem.. From prime.provogue at gmail.com Sat Mar 31 11:33:19 2007 From: prime.provogue at gmail.com (niyaz chistee) Date: Sat, 31 Mar 2007 17:03:19 +0530 Subject: [et-mgmt-tools] Kernel panic while booting rhel4 via pxe In-Reply-To: <460E4335.5060503@netmagicsolutions.com> References: <3293c1d10703310315n2669ff34lbe681801554ab116@mail.gmail.com> <460E3A6B.3090702@netmagicsolutions.com> <3293c1d10703310351i3d0b45dbxa42124be3ddee136@mail.gmail.com> <460E4335.5060503@netmagicsolutions.com> Message-ID: <3293c1d10703310433h38b2f7bncc395f08f3187057@mail.gmail.com> Thanks for the quick reply once again. I did append ramdisk_size=8192 or ramdisk_size=16438 but it didn't helped. Same error once again. One more thing to tell you is that I am trying to install the client via VMware in windows. Does this situation creates this error? Thanks in advance. Regards SysAdmin On 3/31/07, Dhawal Doshy wrote: > > niyaz chistee wrote: > > Thanks for quick response. This is my cobbler --report: > > distro : rhel4 > > kernel : /media/hda6/rhel4/images/pxeboot/vmlinuz > > initrd : /media/hda6/rhel4/images/pxeboot/initrd.img > > kernel options : {} > > architecture : x86 > > ks metadata : {} > > breed : redhat > > > > profile : rhel4x > > distro : rhel4 > > kickstart : /media/hda6/rhel4/ks.cfg > > kernel options : {} > > ks metadata : {} > > virt file size : 5 > > virt ram : 512 > > virt paravirt : True > > repos : > > > > And here is my /tftpboot/pxelinux.cfg/default: > > > > DEFAULT local > > PROMPT 1 > > MENU TITLE Cobbler | http://cobbler.et.redhat.com > > TIMEOUT 200 > > TOTALTIMEOUT 6000 > > ONTIMEOUT local > > > > LABEL local > > MENU LABEL (local) > > MENU DEFAULT > > LOCALBOOT 0 > > > > LABEL rhel4x > > kernel /images/rhel4/vmlinuz > > MENU LABEL rhel4x > > append ksdevice=eth0 lang= syslog=192.168.1.45:25150 > > text initrd=/images/rhel4/initrd.img ks= > > http://192.168.1.45/cblr/kickstarts/rhel4x/ks.cfg > > You do not appear to have a ramdisk_size=8192 OR ramdisk_size=16438 in > your 'append' line of '/tftpboot/pxelinux.cfg/default'. Maybe (just > maybe) that is the cause of the this problem.. > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > -------------- next part -------------- An HTML attachment was scrubbed... URL: