From tobert at gmail.com Sat Dec 1 01:13:40 2007 From: tobert at gmail.com (Al Tobey) Date: Fri, 30 Nov 2007 17:13:40 -0800 Subject: [et-mgmt-tools] koan-0.6.3-3.el4.mrh - kix templating - network rebuilt issue In-Reply-To: <474FEC78.4040506@ng23.net> References: <474B2F75.30008@ng23.net> <474B3C3E.9070708@ng23.net> <474B3DD0.10507@redhat.com> <474B4346.7020506@ng23.net> <474B4401.2030302@redhat.com> <474B49D8.1000208@ng23.net> <474B4CD2.8020804@redhat.com> <474B4F45.9030005@ng23.net> <474F4640.3000008@redhat.com> <474FEC78.4040506@ng23.net> Message-ID: <5ac7acb10711301713o3653aa63n99138713ce29adc7@mail.gmail.com> On Nov 30, 2007 2:56 AM, Tom Brown wrote: > > > > > > > I've applied Al Tobey's --autonet patch to koan upstream. > > > > This seems like it will fix your problem for profile based reinstalls, > > because it will add in the IP info on the kernel command line based on > > what you are currently using. > > > > Might be worth checking out. You should be able to use koan 0.6.4 > > from git with your existing Cobbler install. > > > > thanks for the effort here - alas it does not work for me as it makes > grub look like this > > kernel /vmlinuz ro root=LABEL=/1 ksdevice=eth0 lang= kssendmac syslog=19 > 2.168.10.4:25150 text ks=file:ks.cfg ip=$ip_address netmask=$subnet > hostname=$ho > stname nameserver=$nameserver gateway=$gateway I would think that would cause anaconda to fail anyways. In your template, try putting a check around the net config something like: #if $varExists('ip_address') network --bootproto=static ... #end if -Al > So its not getting the values, its getting the info from the ks template > - Is there no way to get these values from the currently running OS ? > > thanks > > > _______________________________________________ > 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 Sat Dec 1 17:10:44 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Sat, 1 Dec 2007 17:10:44 +0000 Subject: [et-mgmt-tools] Virt-Manager: Supporting additional para-virtual OS's In-Reply-To: <20071130145708.GB10398@redhat.com> References: <473BFA57.7070601@the-nelsons.org> <20071115141911.GF17977@redhat.com> <473CA2BD.5020604@the-nelsons.org> <20071116110215.GB3283@redhat.com> <47467C12.7020602@the-nelsons.org> <20071129204202.GD21587@redhat.com> <474F8E3A.6060709@the-nelsons.org> <20071130145708.GB10398@redhat.com> Message-ID: <20071201171044.GE14465@redhat.com> On Fri, Nov 30, 2007 at 02:57:16PM +0000, Daniel P. Berrange wrote: > On Thu, Nov 29, 2007 at 08:14:50PM -0800, Robert Nelson wrote: > > Daniel P. Berrange wrote: > > >>I spent some time reworking the virtinst code to support OpenSolaris and > > >>make it much easier to support additional OSes. I've attached the patch > > >>file to get some feedback on the work so far. > > >> > > > > > >Thanks for this. Took me a little while to get to grips with it, but I > > >like the resulting structure. I'm going todo a few tests and if it > > >works I'll commit it to the repo. > > > > > >BTW, for any large patches in the future, its handy for review if you > > >do separate patches for re-factoring existing code, vs adding new > > >capabilities. It would have made it easier to review if the splitting > > >up of the DistroManager.py file were a simple no-functional-change > > >refactoring. Also helps people browsing the SCM history in the future > > >to distinguish the changes. No need to change this existing patch > > >though, I'll just add as is... > > > > Please hold off on committing this patch. I have an updated and better > > version coming shortly. I'll post it as two patches to make the SCM > > history more readable. I'll also post it to a new thread. > > So it turned out to be a little more complicated than I expected. At first > I liked the idea of having OSDistro create & use the ImageFetcher class > directly, however, there are circumstances where we need to use the OSDistro > class without the ImageFetcher, so this dependancy causes problems. > > Second, the FullVirtGuest/ParaVirtGuest classes now call into the installer > class to get the disk target - this is a problem because this code is only > relevant for the DistroInstaller class - the ImageInstaller works in a > different way. > > Really, theFullVirt/ParaVirt guest class should not try todo any assignement > of disk targets at all. The 'prepare' method fo the DistroInstaller class > should contain the code to assign disk targets directly. This simplifies > the get_disk_xml methods in the Full/ParaVirtGuest classes, and avoids > the hard dependancy. > > So, in the end I just committed two small parts. I pulled the ImageFetcher > classes out into a separate python module, and pulled the OSDistro > classes out into a separate python module. So just 2 no-functional-change > refactorings for now. I've also just commited the 'LocalImageFetcher' and the modifications to the Guest.py class to catch exceptions when looking up guests for the disk check since they're both worthwhile changes. 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 rjones at redhat.com Tue Dec 4 10:54:26 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Tue, 04 Dec 2007 10:54:26 +0000 Subject: [et-mgmt-tools] solaris virt-install patches In-Reply-To: <521a4d120711300813q60f4b5dfn753b988e9efac255@mail.gmail.com> References: <521a4d120711300813q60f4b5dfn753b988e9efac255@mail.gmail.com> Message-ID: <475531E2.60804@redhat.com> Mark Johnson wrote: > from conversation on #xen.... current Solaris virt-install patches attached. > > Disclaimers :-) > This are informational only (e.g. not proposing they be applied).. > They still need cleanup etc.. These patches are what are in the > current solaris builds. > We are currently still working on the lastest version of virt-install. Hi Mark, It was a bit hard to read your patches because they weren't attached with a text MIME-type. Renaming them as '*.txt' or '*.patch', or reconfiguring your MUA should fix that. mac-address-parse: Looks fine use-virsh-console: Sensible patch outgoing: This patch seems to be corrupt in some way. A diff of a diff? solaris-support: Solaris puts the Python interpreter in /bin instead of /usr/bin. This will break things on Linux so we should fix this so it can find the Python interpreter along the path. No idea how to fix this with Python but I'm sure there's a standard way involving 'env' or something similar. The rest of the patch looks sensible. In general all this code needs rewriting to support capabilities and the storage API (once that's in libvirt), but for a fix now it's fine. Rich. -- Emerging Technologies, Red Hat - http://et.redhat.com/~rjones/ Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SL4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 03798903 -------------- 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 johnson.nh at gmail.com Tue Dec 4 13:05:05 2007 From: johnson.nh at gmail.com (Mark Johnson) Date: Tue, 4 Dec 2007 08:05:05 -0500 Subject: [et-mgmt-tools] solaris virt-install patches In-Reply-To: <475531E2.60804@redhat.com> References: <521a4d120711300813q60f4b5dfn753b988e9efac255@mail.gmail.com> <475531E2.60804@redhat.com> Message-ID: <521a4d120712040505r10c18b79wfc63aa4559f535db@mail.gmail.com> On Dec 4, 2007 5:54 AM, Richard W.M. Jones wrote: > Mark Johnson wrote: > > from conversation on #xen.... current Solaris virt-install patches attached. > > > > Disclaimers :-) > > This are informational only (e.g. not proposing they be applied).. > > They still need cleanup etc.. These patches are what are in the > > current solaris builds. > > We are currently still working on the lastest version of virt-install. > > Hi Mark, > > It was a bit hard to read your patches because they weren't attached > with a text MIME-type. Renaming them as '*.txt' or '*.patch', or > reconfiguring your MUA should fix that. Sorry about that.. :-( > mac-address-parse: Looks fine > > use-virsh-console: Sensible patch > > outgoing: > This patch seems to be corrupt in some way. A diff of a diff? I was in a rush, included that file by mistake.. > solaris-support: > > Solaris puts the Python interpreter in /bin instead of /usr/bin. This > will break things on Linux so we should fix this so it can find the > Python interpreter along the path. No idea how to fix this with Python > but I'm sure there's a standard way involving 'env' or something similar. No need to. We shouldn't be changing that. It's a bug on our end (/bin can be a sym link to /usr/bin). > The rest of the patch looks sensible. In general all this code needs > rewriting to support capabilities and the storage API (once that's in > libvirt), but for a fix now it's fine. Thanks, MRJ From rainer at ultra-secure.de Tue Dec 4 17:58:15 2007 From: rainer at ultra-secure.de (Rainer Duffner) Date: Tue, 04 Dec 2007 18:58:15 +0100 Subject: [et-mgmt-tools] Question about cobbler - variable substitution Message-ID: <47559537.6030506@ultra-secure.de> Hi, I'm doing some tests with cobbler. I created a kickstart-template with things like: network --bootproto=static --device=$ksmetadev --onboot=on --ip=$ksip --netmask=$ksnetmask --gateway=$ksgw --nameserver=$ksns clearpart --all --drives=sda part /boot --fstype ext3 --size=100 --ondisk=sda part pv.2 --size=0 --grow --ondisk=sda volgroup VolGroup00 --pesize=32768 pv.2 logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=512 --grow --maxsize=$ksmaxswap logvol / --fstype ext3 --name=root --vgname=VolGroup00 --size=$ksmaxslash logvol /var/log --fstype ext3 --name=varlog --vgname=VolGroup00 --size=$ksmaxvarlog logvol /var/spool --fstype ext3 --name=varspool --vgname=VolGroup00 --size=$ksmaxvarspool --grow In the "system" config in the webinterface, I have a corresponding ksmeta entry: ksgw=192.168.1.1 ksmaxslash=5000 ksmaxvarspool=2048 ksip=192.168.1.3 ksmetadev=eth0 ksnetmask=255.255.255.0 ksmaxswap=1100 kshostname=host1.intern ksmaxvarlog=2048 ksns=192.168.1.2 I can see that the kickstart file that is being fetched doesn't have all the expansions, yet. It's fetched from http://server//cblr/kickstarts/RHEL5.1-i386-mailserver/ks.cfg Shouldn't the file rather be fetched from the kickstats_sys subdirectory, where a seemingly correctly built kickstart-file is sitting? ksvalidator clears the file from the kickstarts_sys subdirectory (but not, obviously, from the "kickstarts" subdirectory. Can somebody explain this? cheers, Rainer From mdehaan at redhat.com Tue Dec 4 18:42:44 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 04 Dec 2007 13:42:44 -0500 Subject: [et-mgmt-tools] Question about cobbler - variable substitution In-Reply-To: <47559537.6030506@ultra-secure.de> References: <47559537.6030506@ultra-secure.de> Message-ID: <47559FA4.9010801@redhat.com> Rainer Duffner wrote: > Hi, > > I'm doing some tests with cobbler. > I created a kickstart-template with things like: > > network --bootproto=static --device=$ksmetadev --onboot=on --ip=$ksip > --netmask=$ksnetmask --gateway=$ksgw --nameserver=$ksns > > clearpart --all --drives=sda > part /boot --fstype ext3 --size=100 --ondisk=sda > part pv.2 --size=0 --grow --ondisk=sda > volgroup VolGroup00 --pesize=32768 pv.2 > logvol swap --fstype swap --name=LogVol01 --vgname=VolGroup00 --size=512 > --grow --maxsize=$ksmaxswap > logvol / --fstype ext3 --name=root --vgname=VolGroup00 --size=$ksmaxslash > logvol /var/log --fstype ext3 --name=varlog --vgname=VolGroup00 > --size=$ksmaxvarlog > logvol /var/spool --fstype ext3 --name=varspool --vgname=VolGroup00 > --size=$ksmaxvarspool --grow > > > In the "system" config in the webinterface, I have a corresponding > ksmeta entry: > ksgw=192.168.1.1 ksmaxslash=5000 ksmaxvarspool=2048 ksip=192.168.1.3 > ksmetadev=eth0 ksnetmask=255.255.255.0 ksmaxswap=1100 > kshostname=host1.intern ksmaxvarlog=2048 ksns=192.168.1.2 > > I can see that the kickstart file that is being fetched doesn't have all > the expansions, yet. > > It's fetched from > http://server//cblr/kickstarts/RHEL5.1-i386-mailserver/ks.cfg > How are you installing? PXE menus? Koan with --profile? These use the kickstart rendering. PXE with a set MAC address? Koan with --system? These use the kickstarts_sys rendering. From rainer at ultra-secure.de Tue Dec 4 19:55:33 2007 From: rainer at ultra-secure.de (Rainer Duffner) Date: Tue, 4 Dec 2007 20:55:33 +0100 Subject: [et-mgmt-tools] Question about cobbler - variable substitution In-Reply-To: <47559FA4.9010801@redhat.com> References: <47559537.6030506@ultra-secure.de> <47559FA4.9010801@redhat.com> Message-ID: <1F4A16F4-1959-44BD-9544-F064E11D67A9@ultra-secure.de> Am 04.12.2007 um 19:42 schrieb Michael DeHaan: > > How are you installing? > > PXE menus? Koan with --profile? These use the kickstart > rendering. > > PXE with a set MAC address? Koan with --system? These use the > kickstarts_sys rendering. > Well, when it PXE-TFTP-boots, I get into a menu - but I actually don't want to do that ;-) I want to use a MAC-based system. Did I do something wrong? I just added the systems via the web-GUI and assumed it would select the right system via the MAC-address I entered. cheers, Rainer -- Rainer Duffner CISSP, LPI, MCSE rainer at ultra-secure.de From mdehaan at redhat.com Tue Dec 4 20:05:50 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 04 Dec 2007 15:05:50 -0500 Subject: [et-mgmt-tools] Question about cobbler - variable substitution In-Reply-To: <1F4A16F4-1959-44BD-9544-F064E11D67A9@ultra-secure.de> References: <47559537.6030506@ultra-secure.de> <47559FA4.9010801@redhat.com> <1F4A16F4-1959-44BD-9544-F064E11D67A9@ultra-secure.de> Message-ID: <4755B31E.1000307@redhat.com> Rainer Duffner wrote: > > Am 04.12.2007 um 19:42 schrieb Michael DeHaan: > >> >> How are you installing? >> >> PXE menus? Koan with --profile? These use the kickstart rendering. >> >> PXE with a set MAC address? Koan with --system? These use the >> kickstarts_sys rendering. >> > > > Well, when it PXE-TFTP-boots, I get into a menu - but I actually don't > want to do that ;-) > I want to use a MAC-based system. > Did I do something wrong? I just added the systems via the web-GUI and > assumed it would select the right system via the MAC-address I entered. > It will ... provided the MAC addresses match up. It sounds like they might not. If you have the MAC entry, it should boot directly to whatever the system is assigned to and not sit on the menu. From rainer at ultra-secure.de Tue Dec 4 20:23:56 2007 From: rainer at ultra-secure.de (Rainer Duffner) Date: Tue, 4 Dec 2007 21:23:56 +0100 Subject: [et-mgmt-tools] Question about cobbler - variable substitution In-Reply-To: <4755B31E.1000307@redhat.com> References: <47559537.6030506@ultra-secure.de> <47559FA4.9010801@redhat.com> <1F4A16F4-1959-44BD-9544-F064E11D67A9@ultra-secure.de> <4755B31E.1000307@redhat.com> Message-ID: <33B90663-7747-43E2-A840-86F09DEAA20C@ultra-secure.de> Am 04.12.2007 um 21:05 schrieb Michael DeHaan: > Rainer Duffner wrote: >> >> >> Well, when it PXE-TFTP-boots, I get into a menu - but I actually >> don't want to do that ;-) >> I want to use a MAC-based system. >> Did I do something wrong? I just added the systems via the web-GUI >> and assumed it would select the right system via the MAC-address I >> entered. >> > It will ... provided the MAC addresses match up. It sounds like > they might not. > > If you have the MAC entry, it should boot directly to whatever the > system is assigned to and not sit on the menu. > Well, the menu comes after I type "menu". This was already the case in my test-environment, but it generally worked there because I was hardcoding the partition-information and the network-configuration was at the profile-level (which is a mistake, of course). Here's from the "systems" file in /var/lib/cobbler/ : - depth: 2 interfaces: intf0: dhcp_tag: '' gateway: '' hostname: '' ip_address: '' mac_address: '' subnet: '' virt_bridge: '' intf1: dhcp_tag: '' gateway: '' hostname: host143.intern ip_address: '' mac_address: '00:50:56:87:1E:F4' subnet: '' virt_bridge: '' kernel_options: ksdevice: eth1 kickstart: <> ks_meta: ksgw: '192.168.83.129' kshostname: host143.intern ksip: '192.168.83.143' ksmaxslash: '5000' ksmaxswap: '1100' ksmaxvarlog: '2048' ksmaxvarspool: '2048' ksmetadev: eth0 ksnetmask: '255.255.255.128' ksns: '192.168.83.140' name: host143 netboot_enabled: True parent: '' profile: RHEL5.1-i386-mailserver server: <> virt_path: <> virt_type: <> I assume I don't have to assign a fixed IP to every host on the PXE- network via dhcpd.conf? cheers, Rainer -- Rainer Duffner CISSP, LPI, MCSE rainer at ultra-secure.de From mdehaan at redhat.com Tue Dec 4 21:25:13 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 04 Dec 2007 16:25:13 -0500 Subject: [et-mgmt-tools] Question about cobbler - variable substitution In-Reply-To: <33B90663-7747-43E2-A840-86F09DEAA20C@ultra-secure.de> References: <47559537.6030506@ultra-secure.de> <47559FA4.9010801@redhat.com> <1F4A16F4-1959-44BD-9544-F064E11D67A9@ultra-secure.de> <4755B31E.1000307@redhat.com> <33B90663-7747-43E2-A840-86F09DEAA20C@ultra-secure.de> Message-ID: <4755C5B9.8030609@redhat.com> Rainer Duffner wrote: Take a look in /tftpboot/pxelinux.cfg/ and see if you can find the "encoded" file that matches the MAC address you entered. It should be there. This is the file that the PXE boot process will request and use for system configuration. Next, compare that with what you see in the system BIOS to ensure that it actually matches up. It sounds like it's not matching up to me. > I assume I don't have to assign a fixed IP to every host on the > PXE-network via dhcpd.conf? The TFTP/PXE stage that identifies your system by MAC address does not rely on having a static MAC set. The order is encoded MAC, then encoded IP, then default. Except if you are using IA64 arches (elilo), when you must have the IP -- it doesn't do boot-by-MAC. You can add a -vv in your /etc/xinetd.d/tftp file to see a bit more from TFTP in syslog. Hopefully that helps. It really sounds like the MAC addresses aren't right to me, maybe a typo somewhere? > > > cheers, > Rainer From rainer at ultra-secure.de Tue Dec 4 21:39:33 2007 From: rainer at ultra-secure.de (Rainer Duffner) Date: Tue, 4 Dec 2007 22:39:33 +0100 Subject: [et-mgmt-tools] Question about cobbler - variable substitution In-Reply-To: <4755C5B9.8030609@redhat.com> References: <47559537.6030506@ultra-secure.de> <47559FA4.9010801@redhat.com> <1F4A16F4-1959-44BD-9544-F064E11D67A9@ultra-secure.de> <4755B31E.1000307@redhat.com> <33B90663-7747-43E2-A840-86F09DEAA20C@ultra-secure.de> <4755C5B9.8030609@redhat.com> Message-ID: Am 04.12.2007 um 22:25 schrieb Michael DeHaan: > Rainer Duffner wrote: > > Take a look in /tftpboot/pxelinux.cfg/ and see if you can find the > "encoded" file that matches the MAC address you entered. It should > be there. This is > the file that the PXE boot process will request and use for system > configuration. Next, compare that with what you see in the system > BIOS to ensure that it actually matches up. > It sounds like it's not matching up to me. >> I assume I don't have to assign a fixed IP to every host on the >> PXE-network via dhcpd.conf? > > The TFTP/PXE stage that identifies your system by MAC address does > not rely on having a static MAC set. > The order is encoded MAC, then encoded IP, then default. Except if > you are using IA64 arches (elilo), when you must have the IP -- it > doesn't do boot-by-MAC. > You can add a -vv in your /etc/xinetd.d/tftp file to see a bit more > from TFTP in syslog. > > Hopefully that helps. It really sounds like the MAC addresses > aren't right to me, maybe a typo somewhere? > In the above /tftpboot/pxelinux.cfg directory, there's only one file - "default", which produces said menu. Is there anything else that has to be run in addition to "cobbler sync"? I'm feeling I'm missing something, but I don't know what. Thanks for your help so far. cheers, Rainer -- Rainer Duffner CISSP, LPI, MCSE rainer at ultra-secure.de From mdehaan at redhat.com Tue Dec 4 21:50:58 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 04 Dec 2007 16:50:58 -0500 Subject: [et-mgmt-tools] Question about cobbler - variable substitution In-Reply-To: References: <47559537.6030506@ultra-secure.de> <47559FA4.9010801@redhat.com> <1F4A16F4-1959-44BD-9544-F064E11D67A9@ultra-secure.de> <4755B31E.1000307@redhat.com> <33B90663-7747-43E2-A840-86F09DEAA20C@ultra-secure.de> <4755C5B9.8030609@redhat.com> Message-ID: <4755CBC2.9080706@redhat.com> Rainer Duffner wrote: > > Am 04.12.2007 um 22:25 schrieb Michael DeHaan: > >> Rainer Duffner wrote: >> >> Take a look in /tftpboot/pxelinux.cfg/ and see if you can find the >> "encoded" file that matches the MAC address you entered. It should be >> there. This is >> the file that the PXE boot process will request and use for system >> configuration. Next, compare that with what you see in the system >> BIOS to ensure that it actually matches up. >> It sounds like it's not matching up to me. >>> I assume I don't have to assign a fixed IP to every host on the >>> PXE-network via dhcpd.conf? >> >> The TFTP/PXE stage that identifies your system by MAC address does >> not rely on having a static MAC set. >> The order is encoded MAC, then encoded IP, then default. Except if >> you are using IA64 arches (elilo), when you must have the IP -- it >> doesn't do boot-by-MAC. >> You can add a -vv in your /etc/xinetd.d/tftp file to see a bit more >> from TFTP in syslog. >> >> Hopefully that helps. It really sounds like the MAC addresses aren't >> right to me, maybe a typo somewhere? >> > > > In the above /tftpboot/pxelinux.cfg directory, there's only one file - > "default", which produces said menu. > > Is there anything else that has to be run in addition to "cobbler sync"? > > I'm feeling I'm missing something, but I don't know what. > > Thanks for your help so far. > > > cheers, > Rainer In your systems file, I do not see a MAC set for the first interface, just the second. Any chance it was created with --mac1=value versus --mac=value? I can duplicate your problem with the first interface remains empty. This is most definitely a bug, though probably one that few people would encounter. I'll put in a fix. In the meantime, you can edit that file to add the interface to the 'intf0' part and delete the entirity of 'intf1' part, and it will work fine. If you'd rather not worry about tweaking the config files, the Web UI can also do this. Hope that helps... I'll file a defect on this on the Trac page. --Michael From rainer at ultra-secure.de Tue Dec 4 22:23:02 2007 From: rainer at ultra-secure.de (Rainer Duffner) Date: Tue, 4 Dec 2007 23:23:02 +0100 Subject: [et-mgmt-tools] Question about cobbler - variable substitution In-Reply-To: <4755CBC2.9080706@redhat.com> References: <47559537.6030506@ultra-secure.de> <47559FA4.9010801@redhat.com> <1F4A16F4-1959-44BD-9544-F064E11D67A9@ultra-secure.de> <4755B31E.1000307@redhat.com> <33B90663-7747-43E2-A840-86F09DEAA20C@ultra-secure.de> <4755C5B9.8030609@redhat.com> <4755CBC2.9080706@redhat.com> Message-ID: <97B53B7B-3F41-4BC4-AACA-83D779545B37@ultra-secure.de> Am 04.12.2007 um 22:50 schrieb Michael DeHaan: > > > In your systems file, I do not see a MAC set for the first > interface, just the second. Well, I can put in a MAC - but I thought it was not relevant, because I do PXE-boot on eth1 (the "backend interface"). I didn't know it would generate a pxelinux.cfg file out of that MAC... > > Any chance it was created with --mac1=value versus --mac=value? > It was created via the web-interface. Does the web-interface log the XMLRPC-commands somewhere? > I can duplicate your problem with the first interface remains > empty. This is most definitely > a bug, though probably one that few people would encounter. > > I'll put in a fix. In the meantime, you can edit that file to add > the interface to the 'intf0' part > and delete the entirity of 'intf1' part, and it will work fine. If > you'd rather not worry about > tweaking the config files, the Web UI can also do this. > > Hope that helps... I'll file a defect on this on the Trac page. > So, you mean I could make all the settings that I currently have on eth1 also on eth0 and omit eth1? And ksdevice: eth1 will do the trick? ...tries it.... Ha, now I have another file in that directory. I cannot go further, because the hardware is actually an ESX-guest, and the VI-client only runs on Windoze (which I don't have). I'll work on this tomorrow (GMT+1). Again: thanks a lot for your efforts. cheers, Rainer -- Rainer Duffner CISSP, LPI, MCSE rainer at ultra-secure.de From mdehaan at redhat.com Tue Dec 4 22:53:40 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 04 Dec 2007 17:53:40 -0500 Subject: [et-mgmt-tools] Question about cobbler - variable substitution In-Reply-To: <97B53B7B-3F41-4BC4-AACA-83D779545B37@ultra-secure.de> References: <47559537.6030506@ultra-secure.de> <47559FA4.9010801@redhat.com> <1F4A16F4-1959-44BD-9544-F064E11D67A9@ultra-secure.de> <4755B31E.1000307@redhat.com> <33B90663-7747-43E2-A840-86F09DEAA20C@ultra-secure.de> <4755C5B9.8030609@redhat.com> <4755CBC2.9080706@redhat.com> <97B53B7B-3F41-4BC4-AACA-83D779545B37@ultra-secure.de> Message-ID: <4755DA74.7080106@redhat.com> Rainer Duffner wrote: > > Am 04.12.2007 um 22:50 schrieb Michael DeHaan: > >> >> >> In your systems file, I do not see a MAC set for the first interface, >> just the second. > > > Well, I can put in a MAC - but I thought it was not relevant, because > I do PXE-boot on eth1 (the "backend interface"). > I didn't know it would generate a pxelinux.cfg file out of that MAC... The first interface in cobbler's config is not neccessarily 'eth1'. It's just "the first interface in cobbler". > > >> >> Any chance it was created with --mac1=value versus --mac=value? >> > > > It was created via the web-interface. > Does the web-interface log the XMLRPC-commands somewhere? Not in audit log levels of detail ... yet ... though the command you issued was actually valid. The problem seems to lie in the sync behavior for not generating the second file. > > >> I can duplicate your problem with the first interface remains empty. >> This is most definitely >> a bug, though probably one that few people would encounter. >> >> I'll put in a fix. In the meantime, you can edit that file to add the >> interface to the 'intf0' part >> and delete the entirity of 'intf1' part, and it will work fine. If >> you'd rather not worry about >> tweaking the config files, the Web UI can also do this. >> >> Hope that helps... I'll file a defect on this on the Trac page. >> > > > > So, you mean I could make all the settings that I currently have on > eth1 also on eth0 and omit eth1? > And ksdevice: eth1 will do the trick? Yes. > > ...tries it.... > > Ha, now I have another file in that directory. Good deal! From dlutter at redhat.com Wed Dec 5 00:44:56 2007 From: dlutter at redhat.com (David Lutterkort) Date: Tue, 04 Dec 2007 16:44:56 -0800 Subject: [et-mgmt-tools] [PATCH 1 of 4] Allow relative paths, not just file names where files are referenced In-Reply-To: Message-ID: <4227c03796d643c4e906.1196815496@localhost.localdomain> diff -r 6fb5cb8419d3 -r 4227c03796d6 doc/image.rng --- a/doc/image.rng Sat Dec 01 12:12:41 2007 -0500 +++ b/doc/image.rng Tue Dec 04 15:17:10 2007 -0800 @@ -116,9 +116,9 @@ Boot a kernel/initrd contained in the image --> - - - + + + @@ -135,7 +135,7 @@ --> - + @@ -161,7 +161,7 @@ iso for an ISO --> - + system @@ -229,6 +229,11 @@ [^/]+ + + + [a-zA-Z0-9_\+\-%.][a-zA-Z0-9_\+\-%./]* + + [a-zA-Z0-9_\-:./]+ From dlutter at redhat.com Wed Dec 5 00:44:55 2007 From: dlutter at redhat.com (David Lutterkort) Date: Tue, 04 Dec 2007 16:44:55 -0800 Subject: [et-mgmt-tools] [PATCH 0 of 4] virt-image cleanup and small enhancement Message-ID: These patches clean up some minor bugs/inconsistencies in virt-image and make it possible to create a VM without any interaction, simply by running 'virt-image image.xml' From dlutter at redhat.com Wed Dec 5 00:44:57 2007 From: dlutter at redhat.com (David Lutterkort) Date: Tue, 04 Dec 2007 16:44:57 -0800 Subject: [et-mgmt-tools] [PATCH 2 of 4] Fix typo and better error message In-Reply-To: Message-ID: <11d9412efc345d54141a.1196815497@localhost.localdomain> When generating error message about a bad reference from a drive to a disk, use disk_id attribute for the drive and make it clearer that we are talking about a problem in the XML, not that a file is actually missing. diff -r 4227c03796d6 -r 11d9412efc34 virtinst/ImageParser.py --- a/virtinst/ImageParser.py Tue Dec 04 15:17:10 2007 -0800 +++ b/virtinst/ImageParser.py Tue Dec 04 15:17:12 2007 -0800 @@ -60,8 +60,8 @@ class Image: for boot in self.domain.boots: for d in boot.disks: if not self.storage.has_key(d.disk_id): - raise ParserException(_("Disk '%s' not found") - % d.file) + raise ParserException(_("Disk entry for '%s' not found") + % d.disk_id) d.disk = self.storage[d.disk_id] class Domain: From dlutter at redhat.com Wed Dec 5 00:44:59 2007 From: dlutter at redhat.com (David Lutterkort) Date: Tue, 04 Dec 2007 16:44:59 -0800 Subject: [et-mgmt-tools] [PATCH 4 of 4] Don't ask questions on the command line In-Reply-To: Message-ID: diff -r ea7588e72c26 -r eb09d45157b3 virt-image --- a/virt-image Tue Dec 04 15:17:48 2007 -0800 +++ b/virt-image Tue Dec 04 15:17:50 2007 -0800 @@ -41,6 +41,11 @@ gettext.install(virtinst.gettext_app, vi ### General input gathering functions +def get_name(name, image_name, guest): + if name is None: + name = image_name + cli.get_name(name, guest) + def get_memory(memory, image_memory, guest): if memory is None and image_memory is not None: memory = int(image_memory)/1024 @@ -72,6 +77,8 @@ def get_graphics(domain, vnc, vncport, n guest.graphics = False return else: + if not vnc and not sdl: + vnc = True cli.get_graphics(vnc, vncport, nographics, sdl, keymap, guest) ### Option parsing @@ -186,7 +193,7 @@ def main(): guest = virtinst.ParaVirtGuest(connection=conn, installer=installer) # now let's get some of the common questions out of the way - cli.get_name(options.name, guest) + get_name(options.name, image.name, guest) get_memory(options.memory, image.domain.memory, guest) cli.get_uuid(options.uuid, guest) get_vcpus(options.vcpus, image.domain.vcpu, options.check_cpu, From dlutter at redhat.com Wed Dec 5 00:44:58 2007 From: dlutter at redhat.com (David Lutterkort) Date: Tue, 04 Dec 2007 16:44:58 -0800 Subject: [et-mgmt-tools] [PATCH 3 of 4] Allow naming disk storage elements with an id and use that to reference them from the boot descriptor In-Reply-To: Message-ID: diff -r 11d9412efc34 -r ea7588e72c26 doc/image.rng --- a/doc/image.rng Tue Dec 04 15:17:12 2007 -0800 +++ b/doc/image.rng Tue Dec 04 15:17:48 2007 -0800 @@ -150,8 +150,12 @@ + + + diff -r 11d9412efc34 -r ea7588e72c26 virtinst/ImageParser.py --- a/virtinst/ImageParser.py Tue Dec 04 15:17:12 2007 -0800 +++ b/virtinst/ImageParser.py Tue Dec 04 15:17:48 2007 -0800 @@ -46,11 +46,12 @@ class Image: for d in node.xpathEval("storage/disk"): disk = Disk(d) if disk.file is None: + disk.id = "disk%d.img" % len(self.storage) disk.file = "disk%d.img" % (len(self.storage) + 1) - if self.storage.has_key(disk.file): + if self.storage.has_key(disk.id): raise ParserException("Disk file '%s' defined twice" % disk.file) - self.storage[disk.file] = disk + self.storage[disk.id] = disk lm = node.xpathEval("domain") if len(lm) == 1: self.domain = Domain(lm[0]) @@ -176,6 +177,7 @@ class Disk: USE_SCRATCH = "scratch" def __init__(self, node = None): + self.id = None self.file = None self.format = None self.size = None @@ -185,6 +187,7 @@ class Disk: def parseXML(self, node): self.file = xpathString(node, "@file") + self.id = xpathString(node, "@id", self.file) self.format = xpathString(node, "@format", Disk.FORMAT_RAW) self.size = xpathString(node, "@size") self.use = xpathString(node, "@use", Disk.USE_SYSTEM) From berrange at redhat.com Wed Dec 5 00:51:28 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 5 Dec 2007 00:51:28 +0000 Subject: [et-mgmt-tools] [PATCH 1 of 4] Allow relative paths, not just file names where files are referenced In-Reply-To: <4227c03796d643c4e906.1196815496@localhost.localdomain> References: <4227c03796d643c4e906.1196815496@localhost.localdomain> Message-ID: <20071205005127.GA22561@redhat.com> On Tue, Dec 04, 2007 at 04:44:56PM -0800, David Lutterkort wrote: > diff -r 6fb5cb8419d3 -r 4227c03796d6 doc/image.rng > --- a/doc/image.rng Sat Dec 01 12:12:41 2007 -0500 > +++ b/doc/image.rng Tue Dec 04 15:17:10 2007 -0800 > @@ -116,9 +116,9 @@ > Boot a kernel/initrd contained in the image > --> > > - > - > - > + > + > + > > > > @@ -135,7 +135,7 @@ Is this really just a schema change - no code change required ? In any case, ACK. 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 berrange at redhat.com Wed Dec 5 00:51:52 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 5 Dec 2007 00:51:52 +0000 Subject: [et-mgmt-tools] [PATCH 2 of 4] Fix typo and better error message In-Reply-To: <11d9412efc345d54141a.1196815497@localhost.localdomain> References: <11d9412efc345d54141a.1196815497@localhost.localdomain> Message-ID: <20071205005152.GB22561@redhat.com> On Tue, Dec 04, 2007 at 04:44:57PM -0800, David Lutterkort wrote: > When generating error message about a bad reference from a drive to a disk, > use disk_id attribute for the drive and make it clearer that we are talking > about a problem in the XML, not that a file is actually missing. ACK 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 berrange at redhat.com Wed Dec 5 00:52:39 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 5 Dec 2007 00:52:39 +0000 Subject: [et-mgmt-tools] [PATCH 3 of 4] Allow naming disk storage elements with an id and use that to reference them from the boot descriptor In-Reply-To: References: Message-ID: <20071205005239.GC22561@redhat.com> [s..nip...] ACK. 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 berrange at redhat.com Wed Dec 5 00:53:03 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 5 Dec 2007 00:53:03 +0000 Subject: [et-mgmt-tools] [PATCH 4 of 4] Don't ask questions on the command line In-Reply-To: References: Message-ID: <20071205005303.GD22561@redhat.com> On Tue, Dec 04, 2007 at 04:44:59PM -0800, David Lutterkort wrote: > diff -r ea7588e72c26 -r eb09d45157b3 virt-image > --- a/virt-image Tue Dec 04 15:17:48 2007 -0800 > +++ b/virt-image Tue Dec 04 15:17:50 2007 -0800 > @@ -41,6 +41,11 @@ gettext.install(virtinst.gettext_app, vi > > ### General input gathering functions > > +def get_name(name, image_name, guest): > + if name is None: > + name = image_name > + cli.get_name(name, guest) > + > def get_memory(memory, image_memory, guest): > if memory is None and image_memory is not None: > memory = int(image_memory)/1024 > @@ -72,6 +77,8 @@ def get_graphics(domain, vnc, vncport, n > guest.graphics = False > return > else: > + if not vnc and not sdl: > + vnc = True > cli.get_graphics(vnc, vncport, nographics, sdl, keymap, guest) > > ### Option parsing > @@ -186,7 +193,7 @@ def main(): > guest = virtinst.ParaVirtGuest(connection=conn, installer=installer) > > # now let's get some of the common questions out of the way > - cli.get_name(options.name, guest) > + get_name(options.name, image.name, guest) > get_memory(options.memory, image.domain.memory, guest) > cli.get_uuid(options.uuid, guest) > get_vcpus(options.vcpus, image.domain.vcpu, options.check_cpu, ACK 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 dlutter at redhat.com Wed Dec 5 00:58:07 2007 From: dlutter at redhat.com (David Lutterkort) Date: Tue, 04 Dec 2007 16:58:07 -0800 Subject: [et-mgmt-tools] [PATCH 1 of 4] Allow relative paths, not just file names where files are referenced In-Reply-To: <20071205005127.GA22561@redhat.com> References: <4227c03796d643c4e906.1196815496@localhost.localdomain> <20071205005127.GA22561@redhat.com> Message-ID: <1196816287.9156.23.camel@localhost.localdomain> On Wed, 2007-12-05 at 00:51 +0000, Daniel P. Berrange wrote: > On Tue, Dec 04, 2007 at 04:44:56PM -0800, David Lutterkort wrote: > > diff -r 6fb5cb8419d3 -r 4227c03796d6 doc/image.rng > > --- a/doc/image.rng Sat Dec 01 12:12:41 2007 -0500 > > +++ b/doc/image.rng Tue Dec 04 15:17:10 2007 -0800 > > @@ -116,9 +116,9 @@ > > Boot a kernel/initrd contained in the image > > --> > > > > - > > - > > - > > + > > + > > + > > > > > > > > @@ -135,7 +135,7 @@ > > Is this really just a schema change - no code change required ? In any case, > ACK. Yeah, it is. The code is probably a little too permissive in the file names it accepts. David From berrange at redhat.com Wed Dec 5 01:06:37 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 5 Dec 2007 01:06:37 +0000 Subject: [et-mgmt-tools] solaris virt-install patches In-Reply-To: <521a4d120711300813q60f4b5dfn753b988e9efac255@mail.gmail.com> References: <521a4d120711300813q60f4b5dfn753b988e9efac255@mail.gmail.com> Message-ID: <20071205010637.GE22561@redhat.com> On Fri, Nov 30, 2007 at 11:13:03AM -0500, Mark Johnson wrote: > from conversation on #xen.... current Solaris virt-install patches attached. > > Disclaimers :-) > This are informational only (e.g. not proposing they be applied).. > They still need cleanup etc.. These patches are what are in the > current solaris builds. > We are currently still working on the lastest version of virt-install. I applied the MAC address patch since that was trivial. The xm console stuff we already fixed. Likewise the vncviewer changes should disappear when you re-base, since we now use virt-viewer. The other changes all basically look sane. Some of them are actually needed on Linux too - eg switching to use capabilities API instead of touching /proc or /sys. I'd recommend moving the SunOS_installxargs method to the OSDistro.py file in the SolarisOSDistro class just to keep it all in one place. Simiarly I'd prefer to see the loopback mount/unmount stuff kept in the ImageFetcher module directly rather than in util.py 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 dlutter at redhat.com Wed Dec 5 01:07:04 2007 From: dlutter at redhat.com (David Lutterkort) Date: Tue, 04 Dec 2007 17:07:04 -0800 Subject: [et-mgmt-tools] [PATCH 0 of 4] virt-image cleanup and small enhancement In-Reply-To: References: Message-ID: <1196816824.9156.25.camel@localhost.localdomain> On Tue, 2007-12-04 at 16:44 -0800, David Lutterkort wrote: > These patches clean up some minor bugs/inconsistencies in virt-image and > make it possible to create a VM without any interaction, simply by running > 'virt-image image.xml' Committed From ug at xcast.jp Wed Dec 5 07:32:11 2007 From: ug at xcast.jp (Yuji -UG- Imai) Date: Wed, 5 Dec 2007 16:32:11 +0900 Subject: [et-mgmt-tools] RPM repository of virt-vactory for i686 Message-ID: <38adbecd0712042332t6dee28far78cb9b52bd9726ee@mail.gmail.com> Hi, It's very small thing. I'm just trying virt-factory with "Virt-Factory Installation & Setup Guide". I happen to face there is no repository directory on http://virt-factory.et.redhat.com/download/repo/fc7/stable/ . I found relaease symbol "fc7" may be adjust with "f7", there are the directory with "f7". Is that typo? -- ug From byte at aeon.com.my Wed Dec 5 10:38:47 2007 From: byte at aeon.com.my (Colin Charles) Date: Wed, 05 Dec 2007 16:08:47 +0530 Subject: [et-mgmt-tools] virt-manager in F8, gives Vista a bluescreen Message-ID: <47567FB7.6050000@aeon.com.my> Hi! I'm currently running kernel 2.6.23.1-49.fc8, with kvm-36-7.fc8.x86_64.rpm and qemu-0.9.0-5.fc8.x86_64.rpm I have fired up virt-manager, and tried to install Windows Vista in an image. I gave my VM 1GB of RAM, and 15GB of disk (once I tried with pre-allocating the disk space, the other time I didn't) The moment that Windows would start, I would get a blue screen, with a stop error This prevents me from installing Windows, inside of KVM. Have others seen this happen? I tried installing this in the version of kvm/qemu that is provided on Ubuntu Gutsy Gibbon, and it actually worked. References to the Ubuntu stuff: * http://packages.ubuntu.com/gutsy/misc/kvm * http://packages.ubuntu.com/gutsy/misc/qemu thanks and kind regards -- Colin Charles, http://www.bytebot.net/ "First they ignore you, then they laugh at you, then they fight you, then you win." -- Mohandas Gandhi From Nick.Couchman at seakr.com Wed Dec 5 19:16:59 2007 From: Nick.Couchman at seakr.com (Nick Couchman) Date: Wed, 05 Dec 2007 12:16:59 -0700 Subject: [et-mgmt-tools] Problem Creating VMs with Virt-Manager Message-ID: <475696BB.87A6.0099.1@seakr.com> I'm trying to use virt-manager to create a new VM. The problem is that, since my machine uses VLAN tagging and has several VLAN interfaces, I've had to customize the bridges on the machine. So, I have VLAN interfaces vlan10, vlan100, vlan30, vlan110, etc., and bridges to match - xenbr10, xenbr100, xenbr30, xenbr110, etc. When I try to create a new VM, though, in virt-manager 0.5.2, once I get to the networking tab I see an option for a virtual network (which has nothing listed in the drop-down box) and an option for physical network (which has eth0 listed but grayed out with reason "no bridge exists"). So, I can't create a VM because I can't get past this page!! Could someone help me figure out how to change the listing on this page such that my xenbr interfaces show up? Thanks, Nick Couchman -------------- next part -------------- An HTML attachment was scrubbed... URL: From berrange at redhat.com Wed Dec 5 19:31:42 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 5 Dec 2007 19:31:42 +0000 Subject: [et-mgmt-tools] Problem Creating VMs with Virt-Manager In-Reply-To: <475696BB.87A6.0099.1@seakr.com> References: <475696BB.87A6.0099.1@seakr.com> Message-ID: <20071205193142.GR12556@redhat.com> On Wed, Dec 05, 2007 at 12:16:59PM -0700, Nick Couchman wrote: > I'm trying to use virt-manager to create a new VM. The problem is that, > since my machine uses VLAN tagging and has several VLAN interfaces, I've > had to customize the bridges on the machine. So, I have VLAN interfaces > vlan10, vlan100, vlan30, vlan110, etc., and bridges to match - xenbr10, > xenbr100, xenbr30, xenbr110, etc. When I try to create a new VM, though, > in virt-manager 0.5.2, once I get to the networking tab I see an option > for a virtual network (which has nothing listed in the drop-down box) and > an option for physical network (which has eth0 listed but grayed out with > reason "no bridge exists"). So, I can't create a VM because I can't get > past this page!! Could someone help me figure out how to change the > listing on this page such that my xenbr interfaces show up? I'm afraid virt-manager doesn't support installing VMs connected to VLAN interfaces at this time. I'd like to get support for doing this in the UI but I don't have a network with VLANs so can't test. If anyone wants to contribute patches to make it work I'd welcome them. Its mostly just a problem of figuring out how to enumerate the list of bridges attached to VLAN interfaces. In the meantime the best workaround I can offer is to use virt-install command line for installing the VMs initially. This lets you explicitly enter a bridge name as a command line flag. 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 Nick.Couchman at seakr.com Wed Dec 5 19:39:36 2007 From: Nick.Couchman at seakr.com (Nick Couchman) Date: Wed, 05 Dec 2007 12:39:36 -0700 Subject: [et-mgmt-tools] Problem Creating VMs with Virt-Manager In-Reply-To: <20071205193142.GR12556@redhat.com> References: <475696BB.87A6.0099.1@seakr.com> <20071205193142.GR12556@redhat.com> Message-ID: <47569C08.87A6.0099.1@seakr.com> Can you point me in the direction of the file where this would be implemented (where the current code exists for pulling network info)? >>> On Wed, Dec 5, 2007 at 12:31 PM, "Daniel P. Berrange" wrote: On Wed, Dec 05, 2007 at 12:16:59PM -0700, Nick Couchman wrote: > I'm trying to use virt-manager to create a new VM. The problem is that, > since my machine uses VLAN tagging and has several VLAN interfaces, I've > had to customize the bridges on the machine. So, I have VLAN interfaces > vlan10, vlan100, vlan30, vlan110, etc., and bridges to match - xenbr10, > xenbr100, xenbr30, xenbr110, etc. When I try to create a new VM, though, > in virt-manager 0.5.2, once I get to the networking tab I see an option > for a virtual network (which has nothing listed in the drop-down box) and > an option for physical network (which has eth0 listed but grayed out with > reason "no bridge exists"). So, I can't create a VM because I can't get > past this page!! Could someone help me figure out how to change the > listing on this page such that my xenbr interfaces show up? I'm afraid virt-manager doesn't support installing VMs connected to VLAN interfaces at this time. I'd like to get support for doing this in the UI but I don't have a network with VLANs so can't test. If anyone wants to contribute patches to make it work I'd welcome them. Its mostly just a problem of figuring out how to enumerate the list of bridges attached to VLAN interfaces. In the meantime the best workaround I can offer is to use virt-install command line for installing the VMs initially. This lets you explicitly enter a bridge name as a command line flag. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ ( http://search.cpan.org/~danberr/ ) -=| |=- Projects: http://freshmeat.net/~danielpb/ ( 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 -------------- next part -------------- An HTML attachment was scrubbed... URL: From berrange at redhat.com Wed Dec 5 19:44:23 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 5 Dec 2007 19:44:23 +0000 Subject: [et-mgmt-tools] Problem Creating VMs with Virt-Manager In-Reply-To: <47569C08.87A6.0099.1@seakr.com> References: <475696BB.87A6.0099.1@seakr.com> <20071205193142.GR12556@redhat.com> <47569C08.87A6.0099.1@seakr.com> Message-ID: <20071205194423.GS12556@redhat.com> On Wed, Dec 05, 2007 at 12:39:36PM -0700, Nick Couchman wrote: > Can you point me in the direction of the file where this would be > implemented (where the current code exists for pulling network info)? Its in the detect_network_devices()/_device_added() methods for the file /usr/share/virt-manager/virtManager/connection.py. Basically the create new VM wizard will show any device for which there is a vmmNetDevice() registered. 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 javiplx at gmail.com Wed Dec 5 20:05:44 2007 From: javiplx at gmail.com (Javier Palacios) Date: Wed, 5 Dec 2007 21:05:44 +0100 Subject: [et-mgmt-tools] Installing cobbler on a debian system Message-ID: Hello, I want to install cobbler on a debian etch system (not yet to deploy debian systems). Is there any documentation or advices about it ? Javier Palacios From mdehaan at redhat.com Wed Dec 5 20:51:18 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 05 Dec 2007 15:51:18 -0500 Subject: [et-mgmt-tools] Installing cobbler on a debian system In-Reply-To: References: Message-ID: <47570F46.90805@redhat.com> Javier Palacios wrote: > Hello, > > I want to install cobbler on a debian etch system (not yet to deploy > debian systems). Is there any documentation or advices about it ? > > Javier Palacios > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > Not really (it's been used to deploy Debian systems though, see "cobbler distro add --breed=debian"), but if you run into anything, we can probably work through them. I believe "make install" from a source checkout is largely functional on Debian, though there are some places where the code relies on /sbin/service, createrepo, and some other tools. (It would be much happier on a F8 machine.) --Michael From mdehaan at redhat.com Wed Dec 5 21:21:40 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 05 Dec 2007 16:21:40 -0500 Subject: [et-mgmt-tools] Cobbler gets pluggable authentication/authorization (devel branch) Message-ID: <47571664.2010404@redhat.com> Ok, I've implemented the first bits of a customizable authentication and authorization system in Cobbler (0.7.x branch), that should be adaptable to most complex workflows. In other words, you can now define who gets to log in, in your own way, and who gets to do what -- whether that means kerberos/LDAP (FreeIPA?), htdigest/all access, something built on PolicyKit, or something you have in house. (I still need to write some plugins for some of these -- contributions welcome!). The WebUI also now uses mod_python, which allows us to do some nifty tricks like using the same auth system on the frontend as with the web service. That's perhaps less interesting though... Start of documentation on this here: https://hosted.fedoraproject.org/projects/cobbler/wiki/CustomizableSecurity The main advantage to people who don't care about the above is that WebUI setup is a few steps simpler now: https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebInterface You'll notice some permissions based steps are gone, and there's one less authentication file to set up. The other simple change I want to make is to allow the Web UI to log directly in the Apache error logs, so it will be even easier to tell what's going on. It does some of this directly, but it can log more information, and that's the first place people generally look for web based errors anyway. We've also talked here about having logging also be module-based, so more finer grained logging from the XMLRPC layer and the command line is in the works too, after this gets polished up some more. So Cobbler's growing up... and hopefully this will make it a lot more usable in larger configurations where the idea of a few admins having full access doesn't quite solve your administration problems. If you're just a small installation that doesn't care about this kind of thing, Cobbler will of course not force any of this on you... which is also a good thing. Thoughts welcome. --Michael From fj7716hz at aa.jp.fujitsu.com Thu Dec 6 07:45:04 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Thu, 06 Dec 2007 16:45:04 +0900 Subject: [et-mgmt-tools] [PATCH] fix virt-install error: "global name 'OptionValueError' is not defined" Message-ID: <4757A880.3070502@aa.jp.fujitsu.com> Hi, I try virt-install command as follows, # virt-install --name= virt-install command says "global name 'OptionValueError' is not defined". ERROR:root:global name 'OptionValueError' is not defined Traceback (most recent call last): File "/usr/bin/virt-install", line 499, in ? main() File "/usr/bin/virt-install", line 347, in main options = parse_args() File "/usr/bin/virt-install", line 295, in parse_args (options,args) = parser.parse_args() File "/usr/lib/python2.4/optparse.py", line 1275, in parse_args stop = self._process_args(largs, rargs, values) File "/usr/lib/python2.4/optparse.py", line 1315, in _process_args self._process_long_opt(rargs, values) File "/usr/lib/python2.4/optparse.py", line 1390, in _process_long_opt option.process(opt, value, values, self) File "/usr/lib/python2.4/optparse.py", line 707, in process return self.take_action( File "/usr/lib/python2.4/optparse.py", line 726, in take_action self.callback(self, opt, value, parser, *args, **kwargs) File "/usr/lib/python2.4/site-packages/virtinst/cli.py", line 238, in check_before_store raise OptionValueError, _("%s option requires an argument") %opt_str NameError: global name 'OptionValueError' is not defined So, I make a patch to solve this problem. Signed-off-by: Tatsuro Enokura Thanks, Tatsuro Enokura. ---- diff -r c98f26991bad virtinst/cli.py --- a/virtinst/cli.py Wed Dec 05 15:34:29 2007 -0800 +++ b/virtinst/cli.py Thu Dec 06 13:36:38 2007 +0900 @@ -22,6 +22,7 @@ import os, sys import os, sys import logging import logging.handlers +from optparse import OptionValueError import libvirt import util ---- From fj7716hz at aa.jp.fujitsu.com Thu Dec 6 08:33:39 2007 From: fj7716hz at aa.jp.fujitsu.com (Tatsuro Enokura) Date: Thu, 06 Dec 2007 17:33:39 +0900 Subject: [et-mgmt-tools] [PATCH] BZ#251643: Allow to set the cpu pinning for to create installing domain Message-ID: <4757B3E3.5040005@aa.jp.fujitsu.com> Hi, I want to set the cpu pinning for create installing domain by virt-install on RHEL5.1. So, I just add the option to virt-install. We will be allowed to set "cpus" parameter in configuration file and to set "cpuset" parameter in create installing domain with "cpuset" option by this patch. command-line example: # virt-install --name=domain1 --ram=512 --file=/dev/sdb --cpuset=0,2-4 ... some notes are follows: - cpuset can only contain numeric, ',', or '-' characters. - cpuset's pCPU numbers must be less than pCPUs. - This is an effective feature for to create installing domain and configuration file with libvirt revision >= 1.789. - This is an effective feature for to create installing domain and configuration file with xen hypervisor. Other hypervisor(like kvm, qemu...) ignores this option. - This feature covers cpupin on virt-install of BZ#251643. It will work about other things(weight, cap, ...) in the future. Signed-off-by: Tatsuro Enokura Thanks, Tatsuro Enokura -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: cpuset.patch URL: From swissslinky at gmail.com Thu Dec 6 13:12:48 2007 From: swissslinky at gmail.com (Dan) Date: Thu, 06 Dec 2007 14:12:48 +0100 Subject: [et-mgmt-tools] Cobbler Authen::Simple::Kerberos In-Reply-To: <47571664.2010404@redhat.com> References: <47571664.2010404@redhat.com> Message-ID: <1196946768.5551.20.camel@slinky> Hi guys, All excited, I'm trying to install my freshly built rpm... but it complains that Authen::Simple::Kerberos is missing.. but it's not missing. I just installed it.. twice even. I also tested with use strict; use warnings; use Authen::Simple::Kerberos in a blank Perl script. This doesn't complain. I've looked under site_perl and it all seems to be there... Anybody any ideas? From javiplx at gmail.com Thu Dec 6 13:55:00 2007 From: javiplx at gmail.com (Javier Palacios) Date: Thu, 6 Dec 2007 14:55:00 +0100 Subject: [et-mgmt-tools] Installing cobbler on a debian system In-Reply-To: <47570F46.90805@redhat.com> References: <47570F46.90805@redhat.com> Message-ID: > Not really (it's been used to deploy Debian systems though, see "cobbler > distro add --breed=debian"), but if you run into anything, we can > probably work through them. Most of our machines are debian/ubuntu, so that was the second step. The manpage states "limited experimental support" for debian/suse but for your comments I understand that it works right. The "distro add" arguments resembles those for a syslinux/pxeboot. Is it posible also to import debian packages, manage debian repos and so on? > I believe "make install" from a source checkout is largely functional on > Debian, though there are some places where the code relies on > /sbin/service, createrepo, > and some other tools. Yes, `make install` did work with just a couple of easy requirements. But locates some files (apache config for example) in places where debian does not look for them. It also requires rhpl (easy but not available as official debian package), and other things to come .... Although I'll probably setup a fedora box, I'll keep trying to install & run cobbler from a debian system. If you think it might be valuable, I'll keep you/the list informed about the (expected) progress. Javi > > (It would be much happier on a F8 machine.) > > --Michael > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From javiplx at gmail.com Thu Dec 6 14:00:34 2007 From: javiplx at gmail.com (Javier Palacios) Date: Thu, 6 Dec 2007 15:00:34 +0100 Subject: [et-mgmt-tools] Bug on "cobbler import" Message-ID: If rsync is not installed, cobbler import fails even if we are importing from a local directory. This happens on a FC7 box with standard cobbler, and also with the latest version from git. The output I get is [root at javif7 ~]# cobbler import --mirror=/scratch1/fc8-iso --name=fc8-cica - rsync -a '/scratch1/fc8-iso/' /var/www/cobbler/ks_mirror/fc8-cica --exclude-from=/etc/cobbler/rsync.exclude --progress /bin/sh: rsync: command not found Command failed Javier Palacios From rainer at ultra-secure.de Thu Dec 6 14:15:32 2007 From: rainer at ultra-secure.de (Rainer Duffner) Date: Thu, 06 Dec 2007 15:15:32 +0100 Subject: [et-mgmt-tools] Bug on "cobbler import" In-Reply-To: References: Message-ID: <47580404.6080206@ultra-secure.de> Javier Palacios wrote: > If rsync is not installed, cobbler import fails even if we are > importing from a local directory. This happens on a FC7 box with > standard cobbler, and also with the latest version from git. > > The output I get is > > [root at javif7 ~]# cobbler import --mirror=/scratch1/fc8-iso --name=fc8-cica > - rsync -a '/scratch1/fc8-iso/' /var/www/cobbler/ks_mirror/fc8-cica > --exclude-from=/etc/cobbler/rsync.exclude --progress > /bin/sh: rsync: command not found > Command failed > I haven't looked at the code, but it probably uses rsync for everything. So, the real bug is maybe the missing dependency of the cobbler RPM... Having got to Install rsync on the server is a non-issue, IMO. cheers, Rainer From javiplx at gmail.com Thu Dec 6 14:32:36 2007 From: javiplx at gmail.com (Javier Palacios) Date: Thu, 6 Dec 2007 15:32:36 +0100 Subject: [et-mgmt-tools] Bug on "cobbler import" In-Reply-To: <47580404.6080206@ultra-secure.de> References: <47580404.6080206@ultra-secure.de> Message-ID: > > If rsync is not installed, cobbler import fails even if we are > > importing from a local directory. This happens on a FC7 box with > > standard cobbler, and also with the latest version from git. > > [ ... \ > > > I haven't looked at the code, but it probably uses rsync for everything. > So, the real bug is maybe the missing dependency of the cobbler RPM... > > Having got to Install rsync on the server is a non-issue, IMO. Yes if you have rsync installed everything works fine, and if it is added to the spec, you'll never discover it. But the point is that cobbler calls rsync (and stops) when rsync is not needed. It's not a fatal or blocking bug, and we must even agree on that a fix is not required, but it is a bug anyway. Javi From hyclak at math.ohiou.edu Thu Dec 6 15:01:19 2007 From: hyclak at math.ohiou.edu (Matt Hyclak) Date: Thu, 6 Dec 2007 10:01:19 -0500 Subject: [et-mgmt-tools] Cobbler Authen::Simple::Kerberos In-Reply-To: <1196946768.5551.20.camel@slinky> References: <47571664.2010404@redhat.com> <1196946768.5551.20.camel@slinky> Message-ID: <20071206150119.GC6134@math.ohiou.edu> On Thu, Dec 06, 2007 at 02:12:48PM +0100, Dan enlightened us: > All excited, I'm trying to install my freshly built rpm... but it > complains that Authen::Simple::Kerberos is missing.. but it's not > missing. I just installed it.. twice even. > What's complaining? How did you install it? > I also tested with > > use strict; use warnings; > use Authen::Simple::Kerberos > > in a blank Perl script. This doesn't complain. I've looked under > site_perl and it all seems to be there... > > Anybody any ideas? > My guess is that you've installed it from source/CPAN, so RPM doesn't know about it and is complaining. Matt -- Matt Hyclak Department of Mathematics Department of Social Work Ohio University (740) 593-1263 From mdehaan at redhat.com Thu Dec 6 15:41:58 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 06 Dec 2007 10:41:58 -0500 Subject: [et-mgmt-tools] Cobbler Authen::Simple::Kerberos In-Reply-To: <20071206150119.GC6134@math.ohiou.edu> References: <47571664.2010404@redhat.com> <1196946768.5551.20.camel@slinky> <20071206150119.GC6134@math.ohiou.edu> Message-ID: <47581846.3040008@redhat.com> > My guess is that you've installed it from source/CPAN, so RPM doesn't know > about it and is complaining. > > Matt > > This is all from the development branch and the RPM auto-requiring dependencies. I've disabled the auto-requires RPM voodoo, as you don't need this module unless you're using the kerberos module (which I'm still working on testing, didn't quite get done last night). So, unless using authn_kerberos in /etc/cobbler/modules.conf now, you won't need the module. --Michael From mdehaan at redhat.com Thu Dec 6 15:43:42 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 06 Dec 2007 10:43:42 -0500 Subject: [et-mgmt-tools] Bug on "cobbler import" In-Reply-To: <47580404.6080206@ultra-secure.de> References: <47580404.6080206@ultra-secure.de> Message-ID: <475818AE.4030105@redhat.com> Rainer Duffner wrote: > Javier Palacios wrote: > >> If rsync is not installed, cobbler import fails even if we are >> importing from a local directory. This happens on a FC7 box with >> standard cobbler, and also with the latest version from git. >> >> The output I get is >> >> [root at javif7 ~]# cobbler import --mirror=/scratch1/fc8-iso --name=fc8-cica >> - rsync -a '/scratch1/fc8-iso/' /var/www/cobbler/ks_mirror/fc8-cica >> --exclude-from=/etc/cobbler/rsync.exclude --progress >> /bin/sh: rsync: command not found >> Command failed >> >> > > > I haven't looked at the code, but it probably uses rsync for everything. > So, the real bug is maybe the missing dependency of the cobbler RPM... > > Having got to Install rsync on the server is a non-issue, IMO. > nearly everything :) I'll add the RPM dep for completeness. From mdehaan at redhat.com Thu Dec 6 15:47:28 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 06 Dec 2007 10:47:28 -0500 Subject: [et-mgmt-tools] Installing cobbler on a debian system In-Reply-To: References: <47570F46.90805@redhat.com> Message-ID: <47581990.1040103@redhat.com> Javier Palacios wrote: >> Not really (it's been used to deploy Debian systems though, see "cobbler >> distro add --breed=debian"), but if you run into anything, we can >> probably work through them. >> > > Most of our machines are debian/ubuntu, so that was the second step. > The manpage states "limited experimental support" for debian/suse but > for your comments I understand that it works right. The "distro add" > arguments resembles those for a syslinux/pxeboot. Is it posible also > to import debian packages, manage debian repos and so on? > No support for apt repos, though patches to add them can be applied. > >> I believe "make install" from a source checkout is largely functional on >> Debian, though there are some places where the code relies on >> /sbin/service, createrepo, >> and some other tools. >> > > Yes, `make install` did work with just a couple of easy requirements. > But locates some files (apache config for example) in places where > debian does not look for them. It also requires rhpl (easy but not > available as official debian package), and other things to come .... > rhpl is there to support eventual translation, should we ever want to translate strings. > Although I'll probably setup a fedora box, I'll keep trying to install > & run cobbler from a debian system. If you think it might be valuable, > I'll keep you/the list informed about the (expected) progress. > Sounds good. > Javi > > > > >> (It would be much happier on a F8 machine.) >> >> --Michael >> >> _______________________________________________ >> et-mgmt-tools mailing list >> et-mgmt-tools at redhat.com >> https://www.redhat.com/mailman/listinfo/et-mgmt-tools >> >> > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From swissslinky at gmail.com Thu Dec 6 16:05:48 2007 From: swissslinky at gmail.com (Dan) Date: Thu, 06 Dec 2007 17:05:48 +0100 Subject: [et-mgmt-tools] Cobbler Authen::Simple::Kerberos In-Reply-To: <47581846.3040008@redhat.com> References: <47571664.2010404@redhat.com> <1196946768.5551.20.camel@slinky> <20071206150119.GC6134@math.ohiou.edu> <47581846.3040008@redhat.com> Message-ID: <1196957148.5551.35.camel@slinky> On Thu, 2007-12-06 at 10:41 -0500, Michael DeHaan wrote: > > My guess is that you've installed it from source/CPAN, so RPM doesn't know > > about it and is complaining. > > > > Matt > > > > > This is all from the development branch and the RPM auto-requiring > dependencies. > > I've disabled the auto-requires RPM voodoo, as you don't need this > module unless you're using the kerberos module (which I'm still working > on testing, didn't quite get done last night). > > So, unless using authn_kerberos in /etc/cobbler/modules.conf now, you > won't need the module. That's kinda what I want to do. I've been groking around Authen::KRB5 for another project today so would like to continue this theme ;-) Can I kill the voodoo so that I can install, or should I wait and pull it later? > --Michael > > > > _______________________________________________ > 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 Dec 6 16:22:35 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 06 Dec 2007 11:22:35 -0500 Subject: [et-mgmt-tools] Cobbler Authen::Simple::Kerberos In-Reply-To: <1196957148.5551.35.camel@slinky> References: <47571664.2010404@redhat.com> <1196946768.5551.20.camel@slinky> <20071206150119.GC6134@math.ohiou.edu> <47581846.3040008@redhat.com> <1196957148.5551.35.camel@slinky> Message-ID: <475821CB.2070700@redhat.com> Dan wrote: > On Thu, 2007-12-06 at 10:41 -0500, Michael DeHaan wrote: > >>> My guess is that you've installed it from source/CPAN, so RPM doesn't know >>> about it and is complaining. >>> >>> Matt >>> >>> >>> >> This is all from the development branch and the RPM auto-requiring >> dependencies. >> >> I've disabled the auto-requires RPM voodoo, as you don't need this >> module unless you're using the kerberos module (which I'm still working >> on testing, didn't quite get done last night). >> >> So, unless using authn_kerberos in /etc/cobbler/modules.conf now, you >> won't need the module. >> > > That's kinda what I want to do. I've been groking around Authen::KRB5 for another project today so would like to continue this theme ;-) > Can I kill the voodoo so that I can install, or should I wait and pull it later? > > I've pushed a fix to the RPM that makes things installable, so you're good to go. From mdehaan at redhat.com Thu Dec 6 19:03:13 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 06 Dec 2007 14:03:13 -0500 Subject: [et-mgmt-tools] Fedora FudCON 2008 -- Jan 11th, Raleigh NC Message-ID: <47584771.5020901@redhat.com> Of potential interest to folks: http://barcamp.org/FUDConRaleigh2008 --Michael From berrange at redhat.com Fri Dec 7 03:16:23 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Fri, 7 Dec 2007 03:16:23 +0000 Subject: [et-mgmt-tools] [PATCH] BZ#251643: Allow to set the cpu pinning for to create installing domain In-Reply-To: <4757B3E3.5040005@aa.jp.fujitsu.com> References: <4757B3E3.5040005@aa.jp.fujitsu.com> Message-ID: <20071207031623.GB13085@redhat.com> On Thu, Dec 06, 2007 at 05:33:39PM +0900, Tatsuro Enokura wrote: > Hi, > > I want to set the cpu pinning for create installing domain by virt-install > on RHEL5.1. > > So, I just add the option to virt-install. > We will be allowed to set "cpus" parameter in configuration file > and to set "cpuset" parameter in create installing domain with "cpuset" option > by this patch. > > command-line example: > # virt-install --name=domain1 --ram=512 --file=/dev/sdb --cpuset=0,2-4 ... > > some notes are follows: > - cpuset can only contain numeric, ',', or '-' characters. > - cpuset's pCPU numbers must be less than pCPUs. > - This is an effective feature for to create installing domain and > configuration file with libvirt revision >= 1.789. > - This is an effective feature for to create installing domain and > configuration file with xen hypervisor. > Other hypervisor(like kvm, qemu...) ignores this option. > - This feature covers cpupin on virt-install of BZ#251643. > It will work about other things(weight, cap, ...) in the future. Thanks, this looks good 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 berrange at redhat.com Fri Dec 7 03:22:09 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Fri, 7 Dec 2007 03:22:09 +0000 Subject: [et-mgmt-tools] [PATCH] fix virt-install error: "global name 'OptionValueError' is not defined" In-Reply-To: <4757A880.3070502@aa.jp.fujitsu.com> References: <4757A880.3070502@aa.jp.fujitsu.com> Message-ID: <20071207032209.GC13085@redhat.com> On Thu, Dec 06, 2007 at 04:45:04PM +0900, Tatsuro Enokura wrote: > Hi, > > I try virt-install command as follows, > # virt-install --name= > virt-install command says "global name 'OptionValueError' is not defined". > ERROR:root:global name 'OptionValueError' is not defined > Traceback (most recent call last): > > > So, I make a patch to solve this problem. Thanks, I've committed this. 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 javiplx at gmail.com Fri Dec 7 15:04:04 2007 From: javiplx at gmail.com (Javier Palacios) Date: Fri, 7 Dec 2007 16:04:04 +0100 Subject: [et-mgmt-tools] Installing cobbler on a debian system In-Reply-To: <47581990.1040103@redhat.com> References: <47570F46.90805@redhat.com> <47581990.1040103@redhat.com> Message-ID: > > Most of our machines are debian/ubuntu, so that was the second step. > > The manpage states "limited experimental support" for debian/suse but > > for your comments I understand that it works right. The "distro add" > > arguments resembles those for a syslinux/pxeboot. Is it posible also > > to import debian packages, manage debian repos and so on? > > No support for apt repos, though patches to add them can be applied. So do any patch exists? (I don't care if they are experimental, as right now I just plan make experiments with them) > > Although I'll probably setup a fedora box, I'll keep trying to install > > & run cobbler from a debian system. If you think it might be valuable, > > I'll keep you/the list informed about the (expected) progress. > > Sounds good. Debian packages are available at http://usuarios2.arsystel.com/javierpb/cobbler/ they need some improvements, but they allow to import a Fedora8 DVD into cobbler (no test possible until next week). Although it might due to this non-standard setup, I have noticed that just after install cobbler and before importing anything, only 'list kickstarts' works in the web interface, and any other list action produces a python exception that is shown in the browser. Javier Palacios From bbauman at wgate.com Fri Dec 7 16:39:42 2007 From: bbauman at wgate.com (Bruce Bauman) Date: Fri, 7 Dec 2007 11:39:42 -0500 Subject: [et-mgmt-tools] default ks.cfg for CentOS 5.1 Message-ID: I have imported from DVD using "cobbler import", but it appears that the default ks.cfg file is generated from /etc/cobbler/kickstart_fc6.ks rather than /etc/cobbler/default.ks. Is this the expected behavior? How do I name a template such that it will be used for a CentOS 5.1 kickstart? Thanks. No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.503 / Virus Database: 269.16.15 - Release Date: 12/5/2007 12:00 AM From rainer at ultra-secure.de Fri Dec 7 16:49:21 2007 From: rainer at ultra-secure.de (Rainer Duffner) Date: Fri, 07 Dec 2007 17:49:21 +0100 Subject: [et-mgmt-tools] default ks.cfg for CentOS 5.1 In-Reply-To: References: Message-ID: <47597991.5020702@ultra-secure.de> Bruce Bauman wrote: > I have imported from DVD using "cobbler import", but it appears that the default ks.cfg file is generated from /etc/cobbler/kickstart_fc6.ks rather than /etc/cobbler/default.ks. Is this the expected behavior? How do I name a template such that it will be used for a CentOS 5.1 kickstart? > > What I have found works best is to start by using an anaconda.ks file from a manual install (really). The kickstarts will start showing up (in the web-gui) once you make use of them in a profile. There's no "add kickstart" process. Overall, cobbler is well worth the effort getting it to run. The little time you invest up-front, you will save later multiple times. cheers, Rainer From mdehaan at redhat.com Mon Dec 10 15:27:12 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 10 Dec 2007 10:27:12 -0500 Subject: [et-mgmt-tools] default ks.cfg for CentOS 5.1 In-Reply-To: References: Message-ID: <475D5AD0.7000206@redhat.com> Bruce Bauman wrote: > I have imported from DVD using "cobbler import", but it appears that the default ks.cfg file is generated from /etc/cobbler/kickstart_fc6.ks rather than /etc/cobbler/default.ks. Is this the expected behavior? How do I name a template such that it will be used for a CentOS 5.1 kickstart? > > The above behavior is correct. The nomenclature "_fc6" really means "this OS has support for using yum repositories during installation", which was first added for FC6. Think of it as "fc6 compatible". The easiest thing to do if you don't like what that kickstart file contains is ... cobbler profile edit --name=foo --kickstart=/srv/mystuff/foo.ks (and of course, run cobbler sync to apply that change to any systems that depend on it) --Michael From mdehaan at redhat.com Mon Dec 10 15:28:37 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 10 Dec 2007 10:28:37 -0500 Subject: [et-mgmt-tools] Installing cobbler on a debian system In-Reply-To: References: <47570F46.90805@redhat.com> <47581990.1040103@redhat.com> Message-ID: <475D5B25.7060501@redhat.com> Javier Palacios wrote: >>> Most of our machines are debian/ubuntu, so that was the second step. >>> The manpage states "limited experimental support" for debian/suse but >>> for your comments I understand that it works right. The "distro add" >>> arguments resembles those for a syslinux/pxeboot. Is it posible also >>> to import debian packages, manage debian repos and so on? >>> >> No support for apt repos, though patches to add them can be applied. >> > > So do any patch exists? (I don't care if they are experimental, as > right now I just plan make experiments with them) > None exist. > >>> Although I'll probably setup a fedora box, I'll keep trying to install >>> & run cobbler from a debian system. If you think it might be valuable, >>> I'll keep you/the list informed about the (expected) progress. >>> >> Sounds good. >> > > Debian packages are available at > http://usuarios2.arsystel.com/javierpb/cobbler/ > they need some improvements, but they allow to import a Fedora8 DVD > into cobbler (no test possible until next week). > > Although it might due to this non-standard setup, I have noticed that > just after install cobbler and before importing anything, only 'list > kickstarts' works in the web interface, and any other list action > produces a python exception that is shown in the browser. > Sounds like it's your setup to me. If you want to forward details (off-list) I'll take a look at it. > Javier Palacios > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From tom at ng23.net Mon Dec 10 15:34:56 2007 From: tom at ng23.net (Tom Brown) Date: Mon, 10 Dec 2007 15:34:56 +0000 Subject: [et-mgmt-tools] Moving Cobbler Server Message-ID: <475D5CA0.3040901@ng23.net> Hi If i install cobbler onto another server how can i migrate the installation across ? would copying /etc/cobbler /var/www/cobbler /var/lib/cobbler suffice along with the apache config ? thanks From kanarip at kanarip.com Mon Dec 10 15:51:33 2007 From: kanarip at kanarip.com (Jeroen van Meeuwen) Date: Mon, 10 Dec 2007 16:51:33 +0100 Subject: [et-mgmt-tools] Moving Cobbler Server In-Reply-To: <475D5CA0.3040901@ng23.net> References: <475D5CA0.3040901@ng23.net> Message-ID: <475D6085.6030700@kanarip.com> Tom Brown wrote: > Hi > > If i install cobbler onto another server how can i migrate the > installation across ? would copying > > /etc/cobbler > /var/www/cobbler > /var/lib/cobbler > > suffice along with the apache config ? > > thanks > Mind /tftpboot, as well as any location you might have repositories mirrored to. Kind regards, Jeroen van Meeuwen -kanarip From mdehaan at redhat.com Mon Dec 10 15:56:27 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 10 Dec 2007 10:56:27 -0500 Subject: [et-mgmt-tools] Moving Cobbler Server In-Reply-To: <475D5CA0.3040901@ng23.net> References: <475D5CA0.3040901@ng23.net> Message-ID: <475D61AB.8030805@redhat.com> Tom Brown wrote: > Hi > > If i install cobbler onto another server how can i migrate the > installation across ? would copying > > /etc/cobbler > /var/www/cobbler > /var/lib/cobbler > > suffice along with the apache config ? > > thanks > Good question. I'll make an article about this on the Wiki. Here's what I'd do: Install Cobbler on the new system. rsync /var/lib/cobbler and /var/www/cobbler from the old system verify Apache is configured correctly edit /var/lib/cobbler/settings to correct the server address Assuming your distribution trees all came from "cobbler import" you won't have to worry about where your distro files come from. Otherwise, do a "cobbler distro list" and check. It may be that you have local kernels in the config, as opposed to something that already resides in /var/www/cobbler as the result of an import. If your yum repositories are mirrored from local directories (ex: /srv/mypackages/*.rpm), you'll have to deal with them, but if they are all remote, then no problem. You don't have to worry about the cobbler mirrored copies because they are in /var/www/cobbler. Cobbler will take care of building /tftpboot and everything else out for you when you run cobbler sync. If you hit anything strange, let us know, and I'll expand the Wiki article. > > > _______________________________________________ > 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 Dec 10 16:30:28 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 10 Dec 2007 11:30:28 -0500 Subject: [et-mgmt-tools] Moving Cobbler Server In-Reply-To: <475D61AB.8030805@redhat.com> References: <475D5CA0.3040901@ng23.net> <475D61AB.8030805@redhat.com> Message-ID: <475D69A4.3080509@redhat.com> Michael DeHaan wrote: > Tom Brown wrote: >> Hi >> >> If i install cobbler onto another server how can i migrate the >> installation across ? would copying >> >> /etc/cobbler >> /var/www/cobbler >> /var/lib/cobbler >> >> suffice along with the apache config ? >> >> thanks >> > Good question. I'll make an article about this on the Wiki. Here it is... https://hosted.fedoraproject.org/projects/cobbler/wiki/MovingYourInstallToAnotherBox I added a little bit more than was in my original email. --Michael From tom at ng23.net Mon Dec 10 16:46:28 2007 From: tom at ng23.net (Tom Brown) Date: Mon, 10 Dec 2007 16:46:28 +0000 Subject: [et-mgmt-tools] Moving Cobbler Server In-Reply-To: <475D69A4.3080509@redhat.com> References: <475D5CA0.3040901@ng23.net> <475D61AB.8030805@redhat.com> <475D69A4.3080509@redhat.com> Message-ID: <475D6D64.2030706@ng23.net> > > Here it is... > https://hosted.fedoraproject.org/projects/cobbler/wiki/MovingYourInstallToAnotherBox > > > I added a little bit more than was in my original email. > great thanks From crobinso at redhat.com Mon Dec 10 18:49:18 2007 From: crobinso at redhat.com (Cole Robinson) Date: Mon, 10 Dec 2007 13:49:18 -0500 Subject: [et-mgmt-tools] [PATCH] virt-manager: Add additional info to disk details section Message-ID: <475D8A2E.9030100@redhat.com> The attached patch adds a couple fields to the details section for a disk device in virt-manager: a 'Driver' field and a 'Permissions' field. The driver field is written as "file" or "tap:aio" etc., and the permissions field appears as "Readonly" or "Read/Write", with ", Sharable" tacked on the end if applicable. Thanks, Cole -- Cole Robinson crobinso at redhat.com -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: disk-driver-perms-details-patch URL: From lists at davehatton.it Tue Dec 11 00:52:11 2007 From: lists at davehatton.it (Dave Hatton) Date: Tue, 11 Dec 2007 00:52:11 +0000 Subject: [et-mgmt-tools] Getting koan/LiveCD running on F8 Message-ID: <475DDF3B.9090406@davehatton.it> I had to make some small changes to get koan/LiveCD building under Fediora 8 and then creating a Fedora 8 install under VMware. for base.cfg the livecd tools don't seem the take the --packages flag anymore so I added the some extras to %package and also sfdisk seemed to want all 4 primary partions configured - although that might be down to me building in a VMware target machine ------------------------------------------ cut ------------------------------------------------- diff -U 10 base.cfg.orig base.cfg --- base.cfg.orig 2007-11-22 11:22:49.000000000 +0000 +++ base.cfg 2007-12-06 16:14:59.000000000 +0000 @@ -1,46 +1,46 @@ lang en_US.UTF-8 -keyboard us -timezone US/Eastern +keyboard uk +timezone --utc Europe/London auth --useshadow --enablemd5 selinux --disabled firewall --disabled rootpw --iscrypted \$1\$mF86/UHC\$WvcIcX2t6crBz2onWxyac. services --disable sshd # TODO: how to replace i386 with $basearch # TODO: apparently calling it fedora-dev instead of a-dev makes things # not work. Perhaps it has something to do with the default repos in # /etc/yum.repos.d not getting properly disabled? -repo --name=todos --baseurl=http://download.fedora.redhat.com/pub/fedora/linux/releases/7/Everything/i386/os/ -repo --name=updatez --baseurl=http://download.fedora.redhat.com/pub/fedora/linux/updates/7/i386/ +repo --name=todos --baseurl=http://repo.starground.co.uk/rpms-fedora/releases/8/Everything/i386/os/ +repo --name=updatez --baseurl=http://repo.starground.co.uk/rpms-fedora/updates/8/i386/ repo --name=newkoan --baseurl=file:///tmp/newkoan/ text bootloader --location=mbr install zerombr part / --fstype ext3 --size=1024 --grow --ondisk=/dev/sda --asprimary part swap --size=1027 --ondisk=/dev/sda --asprimary %packages @base #@core @hardware-support file syslinux kernel bash -util-linux +util-linux-ng koan avahi-tools #aspell-* -m17n-db-* -man-pages-* # gimp help is huge -gimp-help # lose the compat stuff -compat* # space sucks @@ -58,20 +58,25 @@ # duplicate functionality -tomboy -pinfo -wget # scanning takes quite a bit of space :/ -xsane -xsane-gimp # while hplip requires pyqt, it has to go -hplip #-*debuginfo # error +policycoreutils +grub +eject +tree + %post cat > /etc/rc.d/init.d/fedora-live << EOF #!/bin/bash # # live: Init script for live image # # chkconfig: 345 99 99 @@ -99,22 +104,24 @@ chkconfig --level 345 atd off chkconfig --level 345 anacron off chkconfig --level 345 readahead_early off chkconfig --level 345 readahead_later off # Stopgap fix for RH #217966; should be fixed in HAL instead touch /media/.hal-mtab # take over a drive to use as temporary space sfdisk /dev/sda -uM << ESFDISK - ,1000 - ; +,1000 +; +; +; ESFDISK mkfs -t ext3 /dev/sda1 # fix fstab/mtab cat >> /etc/fstab << EFSTAB /dev/sda1 /tmp/boot ext3 defaults,noatime 0 0 EFSTAB cat >> /etc/mtab << EMTAB @@ -130,32 +137,32 @@ mknod /dev/mapper/livecd-rw b 8 0 grub-install --root-directory=/tmp/boot/ --no-floppy /dev/sda # need a grub.conf file to run grubby from within koan cat > /tmp/boot/boot/grub/grub.conf << EGRUB # grub.conf default=0 timeout=5 #splashimage=(hd0,0)/boot/grub/splash.xpm.gz hiddenmenu -title spacer (2.6.21-1.3228.fc7) +title spacer () root (hd0,0) kernel /boot/vmlinuz initrd /boot/initrd.img EGRUB # now we're ready to do it for real INSERT_KOAN_ARGS --livecd # once through debugging # eject -# reboot +reboot EOF chmod 755 /etc/rc.d/init.d/fedora-live /sbin/restorecon /etc/rc.d/init.d/fedora-live /sbin/chkconfig --add fedora-live # save a little bit of space at least... rm -f /boot/initrd* ------------------------------------------ cut ------------------------------------------------- and for build.py I commented the --package stuff and corrected the _DISTRO_ stuff to F8 ------------------------------------------ cut ------------------------------------------------- diff -U 10 build.py build1.py --- build.py 2007-11-22 09:24:05.000000000 +0000 +++ build1.py 2007-12-04 08:25:40.000000000 +0000 @@ -25,23 +25,23 @@ # tweak at your own risk basef = open("./base.cfg") base_config = basef.read() basef.close() # not expected to work with anything lower than FC-7 # use i386 for compatibility, still works with x86_64 # this is the LiveCD's OS, not the target install OS -USE_DISTRO_SHORT = "FC-7" -USE_DISTRO = "FC-7-i386" -USE_REPO = "fc7i386extras" +USE_DISTRO_SHORT = "F8" +USE_DISTRO = "F8-i386" +USE_REPO = "f8i386extras" # packages to put on the LiveCD packages = [ "kernel", "bash", "koan", "policycoreutils", "grub", "eject", "tree" ] #======= def main(args): @@ -87,22 +87,22 @@ cfg.write(base_config.replace("INSERT_KOAN_ARGS", "/usr/bin/koan %s" % options.koan)) cfg.close() # ====== cmd = "livecd-creator" cmd = cmd + " --fslabel=koan-live-cd" cmd = cmd + " --config=/tmp/koanlive.cfg" - for x in packages: - cmd = cmd + " --package=%s" % x +# for x in packages: +# cmd = cmd + " --package=%s" % x print "running: %s" % cmd try: os.remove("koan-live-cd.iso") except: print "existing file not removed" subprocess.call(cmd, shell=True) if __name__ == "__main__": ------------------------------------------ cut ------------------------------------------------- Not much - but I hope this helps. Dave Hatton From jeremy at rosengren.org Mon Dec 10 19:47:05 2007 From: jeremy at rosengren.org (Jeremy Rosengren) Date: Mon, 10 Dec 2007 13:47:05 -0600 Subject: [et-mgmt-tools] $yum_config_stanza with updated, but deleted, repos. Message-ID: <475D97B9.7050004@rosengren.org> Hi, First, thanks for creating such a nice tool for provisioning Linux systems! I only recently found cobbler and have been hard at work implementing it in my test lab ever since. I've been working the repo_mirror functionality in my cobbler setup and appear to be missing some functionality that would be helpful. During some Linux installs (particularly CentOS, but I think also Fedora), there are repository definitions in /etc/yum.repos.d which default to "enabled=0". I'd like to replicate something similar in my own mirrored repositories, as I'm currently mirroring repositories that I want to be available, but not enabled by default. I'd like to confirm that this feature really doesn't exist and that I haven't missed it completely. If it really doesn't exist, I'll probably take a stab at adding it (or at the very least generating a trac feature request). Thanks, -- jeremy From mdehaan at redhat.com Mon Dec 10 19:50:12 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 10 Dec 2007 14:50:12 -0500 Subject: [et-mgmt-tools] $yum_config_stanza with updated, but deleted, repos. In-Reply-To: <475D97B9.7050004@rosengren.org> References: <475D97B9.7050004@rosengren.org> Message-ID: <475D9874.60601@redhat.com> Jeremy Rosengren wrote: > Hi, > > First, thanks for creating such a nice tool for provisioning Linux > systems! I only recently found cobbler and have been hard at work > implementing it in my test lab ever since. > > I've been working the repo_mirror functionality in my cobbler setup > and appear to be missing some functionality that would be helpful. > During some Linux installs (particularly CentOS, but I think also > Fedora), there are repository definitions in /etc/yum.repos.d which > default to "enabled=0". I'd like to replicate something similar in my > own mirrored repositories, as I'm currently mirroring repositories > that I want to be available, but not enabled by default. > > I'd like to confirm that this feature really doesn't exist and that I > haven't missed it completely. If it really doesn't exist, I'll > probably take a stab at adding it (or at the very least generating a > trac feature request). All the mirrored repositories are currently enabled=1 when installed on the guests. Additions would be welcome. (maybe cobbler repo add ... --name --enabled-on-guest=1/0)? Patches should go to the 'devel' branch, if you haven't noticed it already ... https://hosted.fedoraproject.org/projects/cobbler/wiki/PatchProcess It looks like this would warrant adding a "enabled_by_default" value to each Cobbler repo object (defaulting to True), and having the action_sync.py code template these out appropriately in the retemplate_yum_repos functions. If this gets to confusing, it's something I could add, though it will get in faster if you add it :) Thanks! --Michael > > Thanks, > > -- jeremy > > _______________________________________________ > 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 Mon Dec 10 19:57:20 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Mon, 10 Dec 2007 19:57:20 +0000 Subject: [et-mgmt-tools] [PATCH] virt-manager: Add additional info to disk details section In-Reply-To: <475D8A2E.9030100@redhat.com> References: <475D8A2E.9030100@redhat.com> Message-ID: <20071210195720.GC1973@redhat.com> On Mon, Dec 10, 2007 at 01:49:18PM -0500, Cole Robinson wrote: > The attached patch adds a couple fields to the details section for a disk > device in virt-manager: a 'Driver' field and a 'Permissions' field. The > driver field is written as "file" or "tap:aio" etc., and the permissions > field appears as "Readonly" or "Read/Write", with ", Sharable" tacked on > the end if applicable. I've no problem adding the readonly/readwrite/sharable information. I don't want the file:/tap:aio/etc stuff shown. Its an underlying implementation detail over which the admin has no control - we automatically pick the driver required according to virt type - and as such its not useful to show and people aren't going to understand what it means. 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 lists at davehatton.it Wed Dec 12 12:05:49 2007 From: lists at davehatton.it (Dave Hatton) Date: Wed, 12 Dec 2007 12:05:49 +0000 Subject: [et-mgmt-tools] koan build and ip address/hostname Message-ID: <475FCE9D.1070408@davehatton.it> I'm using LiveCD/Koan to build a target server against a cobbler server without DHCP If I use cobbler system add --name=koantest --mac=00:0c:29:20:9f:b5 --ip=172.16.35.169 --hostname=koantest --profile=koantest --kickstart=/var/lib/cobbler/kickstarts/koantest.ks Should I expect the build process to automatically set the ip address and hostname of the target server or should include code to handle this using the variables $ip_address and $hostname in my ks.cfg? Thanks Dave Hatton From agx at sigxcpu.org Tue Dec 11 10:13:05 2007 From: agx at sigxcpu.org (Guido Guenther) Date: Tue, 11 Dec 2007 11:13:05 +0100 Subject: [et-mgmt-tools] virtinst: fix virt-clone on sparse files Message-ID: <20071211101304.GA22133@bogon.ms20.nix> Hi, attached patch fixes libvirt on fully sparse files. Otherwise you'll get a Cloning from bla.img to bla2.img ERROR: local variable 'b' referenced before assignment exception because b is uninitialzed (we never didn't write anything). Patch is against 0.300.1. Please apply, -- Guido -------------- next part -------------- A non-text attachment was scrubbed... Name: virt-clone_sparse.diff Type: text/x-diff Size: 482 bytes Desc: not available URL: From agx at sigxcpu.org Tue Dec 11 13:24:27 2007 From: agx at sigxcpu.org (Guido Guenther) Date: Tue, 11 Dec 2007 14:24:27 +0100 Subject: [et-mgmt-tools] virtinst: add debian flavour Message-ID: <20071211132427.GA337@bogon.ms20.nix> Hi, please add Debian to the list of supported Linx flavous. Patch attached. Cheers, -- Guido -------------- next part -------------- A non-text attachment was scrubbed... Name: debian.diff Type: text/x-diff Size: 990 bytes Desc: not available URL: From berrange at redhat.com Tue Dec 11 16:44:42 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Tue, 11 Dec 2007 16:44:42 +0000 Subject: [et-mgmt-tools] virtinst: fix virt-clone on sparse files In-Reply-To: <20071211101304.GA22133@bogon.ms20.nix> References: <20071211101304.GA22133@bogon.ms20.nix> Message-ID: <20071211164442.GA17457@redhat.com> On Tue, Dec 11, 2007 at 11:13:05AM +0100, Guido Guenther wrote: > Hi, > attached patch fixes libvirt on fully sparse files. Otherwise you'll get a > > Cloning from bla.img to bla2.img > ERROR: local variable 'b' referenced before assignment > > exception because b is uninitialzed (we never didn't write anything). Actually the block referencing 'b' was mis-indented, so I fixed it that way. http://hg.et.redhat.com/virt/applications/virtinst--devel?cs=04bcf7147ee9 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 berrange at redhat.com Tue Dec 11 16:45:54 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Tue, 11 Dec 2007 16:45:54 +0000 Subject: [et-mgmt-tools] virtinst: add debian flavour In-Reply-To: <20071211132427.GA337@bogon.ms20.nix> References: <20071211132427.GA337@bogon.ms20.nix> Message-ID: <20071211164554.GB17457@redhat.com> On Tue, Dec 11, 2007 at 02:24:27PM +0100, Guido Guenther wrote: > Hi, > please add Debian to the list of supported Linx flavous. Patch attached. Thanks I've applied that. http://hg.et.redhat.com/virt/applications/virtinst--devel?cs=81d9288672cb 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 rigg0022 at umn.edu Tue Dec 11 19:16:39 2007 From: rigg0022 at umn.edu (Riggs, Ben) Date: Tue, 11 Dec 2007 13:16:39 -0600 Subject: [et-mgmt-tools] koan build and ip address/hostname In-Reply-To: <475FCE9D.1070408@davehatton.it> References: <475FCE9D.1070408@davehatton.it> Message-ID: <475EE217.7010602@umn.edu> Without DHCP, you need to have the network line in your ks.cfg list the IP address, netmask, gateway, and nameserver. Check out http://fedoraproject.org/wiki/Anaconda/Kickstart for more info. Dave Hatton wrote: > I'm using LiveCD/Koan to build a target server against a cobbler server > without DHCP > > If I use > > cobbler system add --name=koantest --mac=00:0c:29:20:9f:b5 > --ip=172.16.35.169 --hostname=koantest --profile=koantest > --kickstart=/var/lib/cobbler/kickstarts/koantest.ks > > Should I expect the build process to automatically set the ip address > and hostname of the target server or should include code to handle this > using the variables $ip_address and $hostname in my ks.cfg? > > > Thanks > > Dave Hatton > > _______________________________________________ > 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 Thu Dec 13 02:23:58 2007 From: dlutter at redhat.com (David Lutterkort) Date: Wed, 12 Dec 2007 18:23:58 -0800 Subject: [et-mgmt-tools] [PATCH 0 of 5] Package virtual machine images Message-ID: This set of patches (well, really only 5/5) introduces a new 'virt-pack' command that packages a virtual machine image for distribution. To make that a littel more interesting, it also generates the necesary information to run the image under VMWare(tm). The intention is to make it as easy as possible for the budding appliance writer to make their appliance widely available. There are lots of choices for a container for a VM image - and they all suck for various reasons (RPM is useless for Windows users, ZIP on Linux can't handle archive members bigger than 4GB, tar is slightly unfriendly for Windows users) I used tarballs since they seem to have the fewest pitfalls. In the future, another format that looks promising is 7zip[1]; in my testing the compression it performed was very impressive (and took impressively long) The Windows implementation of 7zip can read tarballs, too - in theory; in practice it reads _some_ tarballs, but balks at others, unclear what causes that. The VMWare metadata that is added to the image consists of a VMX file and several VMDK files, one for each disk in the image. The VMX is generally dumb and conservative, but worked in my simplistic testing with VMWare Server on XP. VMDK files are primadonnas, despite VMWares docs; I believe that the ones virt-pack generates do work. None of the innards of the VM are taken into account by virt-pack; if you write appliances, you are well served to include a reasonable set of drivers in the images. Don't cut out support for e1000 or rtl8139 to save a few bytes - those 'appliances' exist and distinguish themselves by running only on VMWare and nowhere else. Anyway, I'd be very interested in any comments/feedback, and even more in any experiences with using this to package VM images and run them under libvirt and/or VMWare(tm). David [1] http://p7zip.sourceforge.net/ From dlutter at redhat.com Thu Dec 13 02:23:59 2007 From: dlutter at redhat.com (David Lutterkort) Date: Wed, 12 Dec 2007 18:23:59 -0800 Subject: [et-mgmt-tools] [PATCH 1 of 5] Add version information to image.xml In-Reply-To: Message-ID: <557938724328c5b26e22.1197512639@localhost.localdomain> # HG changeset patch # User David Lutterkort # Date 1197512462 28800 # Node ID 557938724328c5b26e222f4a79d2e36b3a1e5d80 # Parent 81d9288672cb1e296f4dc09633c8faeee66046b7 Add version information to image.xml diff -r 81d9288672cb -r 557938724328 doc/image.rng --- a/doc/image.rng Tue Dec 11 11:45:06 2007 -0500 +++ b/doc/image.rng Wed Dec 12 18:21:02 2007 -0800 @@ -8,7 +8,12 @@ storage backing the machine --> - + + + + + + @@ -246,4 +251,9 @@ [a-zA-Z0-9_\-:./]+ + + + [0-9\.]+(-[0-9\.]+)? + + diff -r 81d9288672cb -r 557938724328 virtinst/ImageParser.py --- a/virtinst/ImageParser.py Tue Dec 11 11:45:06 2007 -0500 +++ b/virtinst/ImageParser.py Wed Dec 12 18:21:02 2007 -0800 @@ -36,6 +36,8 @@ class Image: self.name = None self.label = None self.descr = None + self.version = None + self.release = None if not node is None: self.parseXML(node) @@ -43,6 +45,8 @@ class Image: self.name = xpathString(node, "name") self.label = xpathString(node, "label") self.descr = xpathString(node, "description") + self.version = xpathString(node, "name/@version") + self.release = xpathString(node, "name/@release") for d in node.xpathEval("storage/disk"): disk = Disk(d) if disk.file is None: From dlutter at redhat.com Thu Dec 13 02:24:00 2007 From: dlutter at redhat.com (David Lutterkort) Date: Wed, 12 Dec 2007 18:24:00 -0800 Subject: [et-mgmt-tools] [PATCH 2 of 5] Move abspath to ImageParser.Image In-Reply-To: Message-ID: <2f713d54e92e7699c43a.1197512640@localhost.localdomain> # HG changeset patch # User David Lutterkort # Date 1197512462 28800 # Node ID 2f713d54e92e7699c43ac828ab4bab326275edd2 # Parent 557938724328c5b26e222f4a79d2e36b3a1e5d80 Move abspath to ImageParser.Image diff -r 557938724328 -r 2f713d54e92e virtinst/ImageManager.py --- a/virtinst/ImageManager.py Wed Dec 12 18:21:02 2007 -0800 +++ b/virtinst/ImageManager.py Wed Dec 12 18:21:02 2007 -0800 @@ -128,7 +128,7 @@ class ImageInstaller(Guest.Installer): return True def _abspath(self, p): - return os.path.abspath(os.path.join(self.image.base, p)) + return self.image.abspath(p) class PlatformMatchException(Exception): def __init__(self, msg): diff -r 557938724328 -r 2f713d54e92e virtinst/ImageParser.py --- a/virtinst/ImageParser.py Wed Dec 12 18:21:02 2007 -0800 +++ b/virtinst/ImageParser.py Wed Dec 12 18:21:02 2007 -0800 @@ -19,6 +19,7 @@ # Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, # MA 02110-1301 USA. +import os.path import libxml2 import CapabilitiesParser from virtinst import _virtinst as _ @@ -40,6 +41,11 @@ class Image: self.release = None if not node is None: self.parseXML(node) + + def abspath(self, p): + """Turn P into an absolute path. Relative paths are taken relative + to self.BASE""" + return os.path.abspath(os.path.join(self.base, p)) def parseXML(self, node): self.name = xpathString(node, "name") From dlutter at redhat.com Thu Dec 13 02:24:01 2007 From: dlutter at redhat.com (David Lutterkort) Date: Wed, 12 Dec 2007 18:24:01 -0800 Subject: [et-mgmt-tools] [PATCH 3 of 5] Store the image filename in the Image class In-Reply-To: Message-ID: # HG changeset patch # User David Lutterkort # Date 1197512462 28800 # Node ID eb8ba74062f4d4f1fcf13578860d049d3e6e4611 # Parent 2f713d54e92e7699c43ac828ab4bab326275edd2 Store the image filename in the Image class diff -r 2f713d54e92e -r eb8ba74062f4 virt-image --- a/virt-image Wed Dec 12 18:21:02 2007 -0800 +++ b/virt-image Wed Dec 12 18:21:02 2007 -0800 @@ -147,17 +147,6 @@ def parse_args(): return options -def parse_image_xml(fname): - if fname is None: - print >> sys.stderr, _("Must provide the location of an image XML file with --image") - sys.exit(1) - if not os.access(fname, os.R_OK): - print >> sys.stderr, _("Can not read %s") % fname - file = open(fname, "r") - xml = file.read() - file.close() - return virtinst.ImageParser.parse(xml, os.path.dirname(fname)) - def main(): options = parse_args() @@ -166,7 +155,7 @@ def main(): conn = cli.getConnection(options.connect) type = None - image = parse_image_xml(options.image) + image = virtinst.ImageParser.parse_file(options.image) capabilities = virtinst.CapabilitiesParser.parse(conn.getCapabilities()) if options.boot is not None: diff -r 2f713d54e92e -r eb8ba74062f4 virtinst/ImageParser.py --- a/virtinst/ImageParser.py Wed Dec 12 18:21:02 2007 -0800 +++ b/virtinst/ImageParser.py Wed Dec 12 18:21:02 2007 -0800 @@ -30,10 +30,17 @@ class ParserException(Exception): class Image: """The toplevel object representing a VM image""" - def __init__(self, node = None, base = "."): + def __init__(self, node = None, base = None, filename = None): self.storage = {} self.domain = None - self.base = base + self.filename = os.path.abspath(filename) + if base is None: + if filename is not None: + self.base = os.path.dirname(filename) + else: + self.base ="." + else: + self.base = base self.name = None self.label = None self.descr = None @@ -217,7 +224,7 @@ def xpathString(node, path, default = No result = default return result -def parse(xml, base): +def parse(xml, filename): """Parse the XML description of a VM image into a data structure. Returns an object of class Image. BASE should be the directory where the disk image files for this image can be found""" @@ -245,10 +252,15 @@ def parse(xml, base): if root.name != "image": raise ParserException(_("Root element is not 'image'")) - image = Image(root) - image.base = base + image = Image(root, filename = filename) finally: doc.freeDoc() return image +def parse_file(filename): + file = open(filename, "r") + xml = file.read() + file.close() + return parse(xml, filename = filename) + From dlutter at redhat.com Thu Dec 13 02:24:02 2007 From: dlutter at redhat.com (David Lutterkort) Date: Wed, 12 Dec 2007 18:24:02 -0800 Subject: [et-mgmt-tools] [PATCH 4 of 5] Rename Boot.disks to Boot.drives In-Reply-To: Message-ID: # HG changeset patch # User David Lutterkort # Date 1197512462 28800 # Node ID da46770a461b6299c409522d82eec58d45e29bf1 # Parent eb8ba74062f4d4f1fcf13578860d049d3e6e4611 Rename Boot.disks to Boot.drives In the XML, the corresponding element is called 'drive', not 'disk' diff -r eb8ba74062f4 -r da46770a461b virtinst/ImageManager.py --- a/virtinst/ImageManager.py Wed Dec 12 18:21:02 2007 -0800 +++ b/virtinst/ImageManager.py Wed Dec 12 18:21:02 2007 -0800 @@ -70,7 +70,7 @@ class ImageInstaller(Guest.Installer): guest.features[f] = False def _make_disks(self, guest): - for m in self.boot.disks: + for m in self.boot.drives: p = self._abspath(m.disk.file) s = None if m.disk.size is not None: diff -r eb8ba74062f4 -r da46770a461b virtinst/ImageParser.py --- a/virtinst/ImageParser.py Wed Dec 12 18:21:02 2007 -0800 +++ b/virtinst/ImageParser.py Wed Dec 12 18:21:02 2007 -0800 @@ -76,7 +76,7 @@ class Image: raise ParserException(_("Expected exactly one 'domain' element")) # Connect the disk maps to the disk definitions for boot in self.domain.boots: - for d in boot.disks: + for d in boot.drives: if not self.storage.has_key(d.disk_id): raise ParserException(_("Disk entry for '%s' not found") % d.disk_id) @@ -135,7 +135,7 @@ class Boot: self.kernel = None self.initrd = None self.cmdline = None - self.disks = [] + self.drives = [] self.arch = None self.features = ImageFeatures() if not node is None: @@ -157,7 +157,7 @@ class Boot: self.features = ImageFeatures(fl[0]) for d in node.xpathEval("drive"): - self.disks.append(Drive(d)) + self.drives.append(Drive(d)) validate(self.type is not None, "The boot type must be provided") From dlutter at redhat.com Thu Dec 13 02:24:03 2007 From: dlutter at redhat.com (David Lutterkort) Date: Wed, 12 Dec 2007 18:24:03 -0800 Subject: [et-mgmt-tools] [PATCH 5 of 5] Package/unpackage an image for distribution In-Reply-To: Message-ID: # HG changeset patch # User David Lutterkort # Date 1197512462 28800 # Node ID e9798e85269935f0a2a92fc964d3794e94c4d32d # Parent da46770a461b6299c409522d82eec58d45e29bf1 Package/unpackage an image for distribution Pack all the files for the image into a tarball. The name and toplevel directory are derived from the name and version of the image descriptor. Scratch disks are omitted, since they can be rebuilt when the image is deployed. To make things more interesting, information for running the image under VMWare is also generated. This consists of a VMX file and VMDK descriptors for each disk in the image. For scratch disks, sparse VMDK files are generated, since VMWare has no notion of scratch disks. The generated VMX file is very conservative, so that it's widely usable, but there are many things that could be improved in it if we had more information about the image (one example: are vmware-tools installed ?) diff -r da46770a461b -r e9798e852699 virt-pack --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/virt-pack Wed Dec 12 18:21:02 2007 -0800 @@ -0,0 +1,149 @@ +#!/usr/bin/python -tt +# +# Package and unpackage images for distribution +# +# Copyright 2007 Red Hat, Inc. +# David Lutterkort +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software +# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, +# MA 02110-1301 USA. + +import os, sys, string +from optparse import OptionParser, OptionValueError +import subprocess +import logging +import libxml2 +import urlgrabber.progress as progress + +import virtinst +import virtinst.ImageParser +import virtinst.CapabilitiesParser +import virtinst.cli as cli +import virtinst.util as util +import virtinst.UnWare + +import tempfile + +import gettext +import locale + +locale.setlocale(locale.LC_ALL, '') +gettext.bindtextdomain(virtinst.gettext_app, virtinst.gettext_dir) +gettext.install(virtinst.gettext_app, virtinst.gettext_dir) + +class PackageException(Exception): + def __init__(self, msg): + Exception.__init__(self, msg) + +class Package: + + def __init__(self, image): + self.image = image + if image.name is None or image.version is None: + raise PackageException( + _("The image name and version must be present")) + self.vername = "%s-%s" % (image.name, image.version) + self.tmpdir = tempfile.mkdtemp(dir="/var/tmp", prefix="virt-pack") + self.stagedir = os.path.join(self.tmpdir, self.vername) + self.files = [] + + def add_image_files(self): + cwd = os.getcwd() + img = self.image + self.files.append(os.path.basename(img.filename)) + try: + os.chdir(img.base) + for d in img.storage.keys(): + disk = img.storage[d] + if disk.use == disk.USE_SCRATCH: + if disk.size is None: + raise PackageException(_("Scratch disk %s does not have a size attribute") % disk.id) + else: + if not os.path.exists(disk.file): + raise PackageException(_("Disk file %s could not be found") % disk.id) + self.files.append(disk.file) + finally: + os.chdir(cwd) + + def make_vmx_files(self): + img = virtinst.UnWare.Image(self.image) + files = img.make(self.image.base) + self.files.extend(files) + + def pack(self, outdir): + outfile = os.path.join(outdir, self.vername + ".tgz") + for f in set(self.files): + dir = os.path.join(self.stagedir, os.path.dirname(f)) + if not os.path.exists(dir): + os.makedirs(dir) + os.symlink(os.path.join(self.image.base, f), + os.path.join(self.stagedir, f)) + print "Writing %s" % outfile + cmd = "tar chzv -C %s -f %s %s" % (self.tmpdir, + outfile, + os.path.basename(self.vername)) + util.system(cmd) + return outfile + +### Option parsing +def parse_args(): + parser = OptionParser() + parser.set_usage("%prog [options] image.xml") + + parser.add_option("-o", "--output", type="string", dest="output", + action="callback", callback=cli.check_before_store, + help=_("Directory in which packaged file will be put")) + parser.add_option("-d", "--debug", action="store_true", dest="debug", + help=_("Print debugging information")) + + (options,args) = parser.parse_args() + if len(args) < 1: + parser.error(_("You need to provide an image XML descriptor")) + options.image = args[0] + + return options + +def main(): + options = parse_args() + + + cli.setupLogging("virt-pack", options.debug) + + image = virtinst.ImageParser.parse_file(options.image) + + if image.name is None or image.version is None: + print >> sys.stderr, _("The image descriptor must contain name and version") + valid = False + + if options.output is None: + options.output = os.path.join(image.base, "..") + + pkg = Package(image) + try: + pkg.add_image_files() + except PackageException, e: + print >> sys.stderr, "Validation failed: %s", e + return 1 + + try: + pkg.make_vmx_files() + pkg.pack(options.output) + except PackageException, e: + print >> sys.stderr, "Packaging failed: %s" % e + return 1 + +if __name__ == "__main__": + main() + diff -r da46770a461b -r e9798e852699 virtinst/UnWare.py --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/virtinst/UnWare.py Wed Dec 12 18:21:02 2007 -0800 @@ -0,0 +1,291 @@ +# +# Processing of VMWare(tm) .vmx files +# +# Copyright 2007 Red Hat, Inc. +# David Lutterkort +# +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. +# +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software +# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, +# MA 02110-1301 USA. + +import time +import sys +import os + +import ImageParser +import util + +class Disk: + """A disk for a VMWare(tm) virtual machine""" + + MONOLITHIC_FLAT = "monolithicFlat" + TWO_GB_MAX_EXTENT_SPARSE = "twoGbMaxExtentSparse" + # This seems only to be usable if the vmdk header is embedded in the + # data file, not when the descriptor is in a separate text file. Use + # TWO_GB_MAX_EXTENT_SPARSE instead. + # VMWare's(tm) documentation of VMDK seriously sucks. A lot. + MONOLITHIC_SPARSE = "monolithicSparse" + + IDE_HEADS = 16 + IDE_SECTORS = 63 + + def __init__(self, descriptor, extent, size, dev, format): + """Create a new disk. DESCRIPTOR is the name of the VMDK descriptor + file. EXTENT is the name of the file holding the actual data. SIZE + is the filesize in bytes. DEV identifies the device, for IDE (the + only one supported right now) it should be $bus:$master. FORMAT is + the format of the underlying extent, one of the formats defined in + virtinst.ImageParser.Disk""" + self.cid = 0xffffffff + self.createType = Disk.MONOLITHIC_FLAT + self.descriptor = descriptor + self.extent = extent + self.size = size + self.dev = dev + self.format = format + + def make_extent(self, base): + """Write the descriptor file, and create the extent as a monolithic + sparse extent if it does not exist yet""" + f = os.path.join(base, self.extent) + print "Checking %s" % f + if not os.path.exists(f): + util.system("qemu-img create -f vmdk %s %d" % (f, self.size/1024)) + self.createType = Disk.TWO_GB_MAX_EXTENT_SPARSE + else: + qemu = os.popen("qemu-img info %s" % f, "r") + for l in qemu: + (tag, val) = l.split(":") + if tag == "file format" and val.strip() == "vmdk": + self.createType = Disk.TWO_GB_MAX_EXTENT_SPARSE + qemu.close() + return self.extent + + def to_vmdk(self): + """Return the VMDK descriptor for this disk""" + + vmdk = """# Disk DescriptorFile +# Generated from libvirt +version=1 +""" + vmdk += "CID=%08x\nparentCID=ffffffff\n" % self.cid + vmdk += "createType=\"%s\"\n\n" % self.createType + vmdk += "# Extent description\n" + blocks = self.size/512 + if self.createType == Disk.MONOLITHIC_FLAT: + vmdk += "RW %d FLAT \"%s\" 0\n" % (blocks, os.path.basename(self.extent)) + else: # Disk.MONOLITHIC_SPARSE + vmdk += "RW %d SPARSE \"%s\"\n" % (blocks, os.path.basename(self.extent)) + + vmdk += """ +# Disk Data Base +ddb.virtualHWVersion = "4" +ddb.adapterType = \"ide\" +ddb.geometry.sectors = \"%d\" +ddb.geometry.heads = \"%d\" +ddb.geometry.cylinders = \"%d\" +""" % (Disk.IDE_SECTORS, Disk.IDE_HEADS, blocks/(Disk.IDE_SECTORS*Disk.IDE_HEADS)) + return vmdk + + def to_vmx(self): + """Return the fragment for the VMX file for this disk""" + + vmx = "" + dict = { + "dev" : self.dev, + "disk_filename" : self.descriptor + } + if self.format == ImageParser.Disk.FORMAT_ISO: + vmx = _VMX_ISO_TEMPLATE % dict + else: # FORMAT_RAW + vmx = _VMX_IDE_TEMPLATE % dict + return vmx + +class Image: + """Represent an image for generation of a VMWare(tm) description""" + + def __init__(self, image = None): + if image is not None: + self._init_from_image(image) + + def _init_from_image(self, image): + domain = image.domain + boot = domain.boots[0] + + self.base = image.base + self.name = image.name + self.descr = image.descr + self.label = image.label + self.vcpu = domain.vcpu + self.memory = domain.memory + self.interface = domain.interface + + self.disks = [] + for d in boot.drives: + disk = d.disk + descriptor = sub_ext(disk.file, ".vmdk") + if disk.size is None: + f = os.path.join(image.base, disk.file) + size = os.stat(f).st_size + else: + size = long(disk.size) * 1024L * 1024L + ide_count = len(self.disks) + dev = "%d:%d" % (ide_count / 2, ide_count % 2) + self.disks.append(Disk(descriptor, disk.file, size, dev, + disk.format)) + + def make(self, base): + """Write the descriptor file and all the disk descriptors""" + files = [] + out = open(os.path.join(self.base, self.name + ".vmx"), "w") + out.write(self.to_vmx()) + out.close() + files.append(self.name + ".vmx") + + for d in self.disks: + f = d.make_extent(self.base) + files.append(f) + out = open(os.path.join(base, d.descriptor), "w") + out.write(d.to_vmdk()) + out.close() + files.append(d.descriptor) + return files + + def to_vmx(self): + """Return the VMX description of this image""" + dict = { + "now": time.strftime("%Y-%m-%dT%H:%M:%S %Z", time.localtime()), + "progname": os.path.basename(sys.argv[0]), + "/image/name": self.name, + "/image/description": self.descr or "None", + "/image/label": self.label or self.name, + "/image/devices/vcpu" : self.vcpu, + "/image/devices/memory": long(self.memory)/1024 + } + + vmx = _VMX_MAIN_TEMPLATE % dict + if self.interface: + vmx += _VMX_ETHER_TEMPLATE + + for d in self.disks: + vmx += d.to_vmx() + + return vmx + +def sub_ext(filename, ext): + return os.path.splitext(filename)[0] + ext + +_VMX_MAIN_TEMPLATE = """ +#!/usr/bin/vmplayer + +# Generated %(now)s by %(progname)s +# http://virt-manager.et.redhat.com/ + +# This is a Workstation 5 or 5.5 config file +# It can be used with Player +config.version = "8" +virtualHW.version = "4" + +# Selected operating system for your virtual machine +guestOS = "other" + +# displayName is your own name for the virtual machine +displayName = "%(/image/name)s" + +# These fields are free text description fields +annotation = "%(/image/description)s" +guestinfo.vmware.product.long = "%(/image/label)s" +guestinfo.vmware.product.url = "http://virt-manager.et.redhat.com/" +guestinfo.vmware.product.class = "virtual machine" + +# Number of virtual CPUs. Your virtual machine will not +# work if this number is higher than the number of your physical CPUs +numvcpus = "%(/image/devices/vcpu)s" + +# Memory size and other memory settings +memsize = "%(/image/devices/memory)d" +MemAllowAutoScaleDown = "FALSE" +MemTrimRate = "-1" + +# Unique ID for the virtual machine will be created +uuid.action = "create" + +## For appliances where tools are installed already, this should really +## be false, but we don't have that ionfo in the metadata +# Remind to install VMware Tools +# This setting has no effect in VMware Player +tools.remindInstall = "TRUE" + +# Startup hints interfers with automatic startup of a virtual machine +# This setting has no effect in VMware Player +hints.hideAll = "TRUE" + +# Enable time synchronization between computer +# and virtual machine +tools.syncTime = "TRUE" + +# First serial port, physical COM1 is not available +serial0.present = "FALSE" + +# Optional second serial port, physical COM2 is not available +serial1.present = "FALSE" + +# First parallell port, physical LPT1 is not available +parallel0.present = "FALSE" + +# Logging +# This config activates logging, and keeps last log +logging = "TRUE" +log.fileName = "%(/image/name)s.log" +log.append = "TRUE" +log.keepOld = "3" + +# These settings decides interaction between your +# computer and the virtual machine +isolation.tools.hgfs.disable = "FALSE" +isolation.tools.dnd.disable = "FALSE" +isolation.tools.copy.enable = "TRUE" +isolation.tools.paste.enabled = "TRUE" + +# Settings for physical floppy drive +floppy0.present = "FALSE" +""" + +_VMX_ETHER_TEMPLATE = """ +## if /image/devices/interface is present: +# First network interface card +ethernet0.present = "TRUE" +ethernet0.connectionType = "nat" +ethernet0.addressType = "generated" +ethernet0.generatedAddressOffset = "0" +ethernet0.autoDetect = "TRUE" +""" + +_VMX_ISO_TEMPLATE = """ +# CDROM drive +ide%(dev)s.present = "TRUE" +ide%(dev)s.deviceType = "cdrom-raw" +ide%(dev)s.startConnected = "TRUE" +ide%(dev)s.fileName = "%(disk_filename)s" +ide%(dev)s.autodetect = "TRUE" +""" + +_VMX_IDE_TEMPLATE = """ +# IDE disk +ide%(dev)s.present = "TRUE" +ide%(dev)s.fileName = "%(disk_filename)s" +ide%(dev)s.mode = "persistent" +ide%(dev)s.startConnected = "TRUE" +ide%(dev)s.writeThrough = "TRUE" +""" diff -r da46770a461b -r e9798e852699 virtinst/util.py --- a/virtinst/util.py Wed Dec 12 18:21:02 2007 -0800 +++ b/virtinst/util.py Wed Dec 12 18:21:02 2007 -0800 @@ -206,3 +206,9 @@ def get_phy_cpus(conn): hostinfo = conn.getInfo() pcpus = hostinfo[4] * hostinfo[5] * hostinfo[6] * hostinfo[7] return pcpus + +def system(cmd): + st = os.system(cmd) + if os.WIFEXITED(st) and os.WEXITSTATUS(st) != 0: + raise OSError("Failed to run %s, exited with %d" % + (cmd, os.WEXITSTATUS(st))) From markmc at redhat.com Thu Dec 13 10:28:33 2007 From: markmc at redhat.com (Mark McLoughlin) Date: Thu, 13 Dec 2007 10:28:33 +0000 Subject: [et-mgmt-tools] [PATCH 5 of 5] Package/unpackage an image for distribution In-Reply-To: References: Message-ID: <1197541713.4428.12.camel@blaa> On Wed, 2007-12-12 at 18:24 -0800, David Lutterkort wrote: > Package/unpackage an image for distribution Interesting ... one thought is that you should expose the Package[1] class, and an UnWarePackage sub-class, from virtinst instead of UnWare. Also, I'd really only expect a single public Package method - pack(). Basically, I'm thinking about how the likes of livecd's image-creator could use this stuff. Cheers, Mark. [1] - Also, I don't think it's usual for constructors to raise exceptions ... I think constructors are always supposed to be safe in python From mdehaan at redhat.com Thu Dec 13 15:57:21 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 13 Dec 2007 10:57:21 -0500 Subject: [et-mgmt-tools] Checkouts for Fedora Hosted Projects (ex: cobbler, koan, etc) Message-ID: <47615661.9070005@redhat.com> Fedora Hosted Projects has changed their addresses for web pages and source repositories around somewhat, so if you have an existing git checkout that is /not/ working, you can fix this by editing your .git/config file and using one of the values listed on the gitweb page for that particular project. For instance, Cobbler's is: http://git.fedoraproject.org/git/cobbler/ git://git.fedorahosted.org/cobbler ssh://git.fedorahosted.org/git/cobbler http://git.fedorahosted.org/git/cobbler For koan, replace "cobbler", with "koan". Simple enough. Existing URLs may continue to work but will probably /not/ work in a month or so. This is all part of Fedora Hosted Projects moving to https://fedorahosted.org/, which should be usable now, but is basically in 'beta' at this point. Feel free to browse all the projects there. If anyone still has problems still getting at source, let me know. --Michael From dlutter at redhat.com Thu Dec 13 17:38:21 2007 From: dlutter at redhat.com (David Lutterkort) Date: Thu, 13 Dec 2007 09:38:21 -0800 Subject: [et-mgmt-tools] [PATCH 5 of 5] Package/unpackage an image for distribution In-Reply-To: <1197541713.4428.12.camel@blaa> References: <1197541713.4428.12.camel@blaa> Message-ID: <1197567501.3124.4.camel@localhost.localdomain> On Thu, 2007-12-13 at 10:28 +0000, Mark McLoughlin wrote: > On Wed, 2007-12-12 at 18:24 -0800, David Lutterkort wrote: > > > Package/unpackage an image for distribution > > Interesting ... one thought is that you should expose the Package[1] > class, and an UnWarePackage sub-class, from virtinst instead of UnWare. > Also, I'd really only expect a single public Package method - pack(). > > Basically, I'm thinking about how the likes of livecd's image-creator > could use this stuff. Yeah, I structured it that way because I didn't think there would be many users for it - how would you include that with image-creator ? Are you thinking of doing the image creation soup-to-nuts (kickstart to tarball, really) in one big jump ? > [1] - Also, I don't think it's usual for constructors to raise > exceptions ... I think constructors are always supposed to be safe in > python Yeah, it's lame .. I'll make it a factory method David From mdehaan at redhat.com Thu Dec 13 23:13:22 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 13 Dec 2007 18:13:22 -0500 Subject: [et-mgmt-tools] Cobbler (0.7.0) and koan (0.6.4) test release builds Message-ID: <4761BC92.3020509@redhat.com> For those interested in what is going on in Cobbler lately, but not following the latest git and/or IRC, I've built some testing RPMs... http://cobbler.et.redhat.com/download/testing/ To summarize, the main new thing in Cobbler is the start of a pluggable authn/authz system and moving the WebUI to mod_python. This means for a faster experience, but also one that can be more easily integrated with other corporate environments. This will evolve more over the next few weeks. Perhaps even more importantly, Cobbler Web UI setup is also much simpler as there are less steps. There's also support for deploying Xen fullvirt (if you have F8 or higher), and logging across the Web services APIs (and in general) are much improved. See https://fedorahosted.org/cobbler/wiki/CobblerWebInterface Koan gets support for Xen fullvirt, diskless Xen systems, and the aforementioned --autonet post to the list. Koan also gets a --no-gfx option back (for Xen installs when you don't want to use VNC) that shouldn't have been removed previously. I plan to continue the 0.7.X branch for another month or so, with more features in the area of performance, authn/authz, and so forth. Feel free to pound on what is there now -- there are likely a few bugs and that always makes for a better release, but feedback and RFEs are always great too. Cobbler 0.6.X will probably not see an additional release unless major bugs come up that we would like to fix. Changelogs are below: Cobbler 0.7.0 - Testing branch - Fix bug related to <> and kickstart args - Make CLI functions modular and use optparse - Quote wget args to avoid creating stray files on target system - Support Xen FV as virt type (requires F8+) - Implemented fully pluggable authn/authz system - WebUI is now mod_python based - Greatly enhanced logging (goes to /var/log/cobbler/cobbler.log) - New --no-triggers and --no-sync on "adds" for performance and other reasons - performance testing scripts (in source checkout) - webui now uses Apache logging - misc webui fixes - ... Koan 0.6.4* * Tue Nov 27 2007 Michael DeHaan - 0.6.4-1 - Added --nogfx option, which is basically the inverse of the old --virt-graphics - Allow Xen machines to be constructed with no disks (size=0), but not qemu/KVM - Support Xen Fullvirt via PXE - Add new --autonet command - Fail if kernel argument length exceeds the 255 cap, as truncation can lead to hard to debug errors. --Michael From swissslinky at gmail.com Fri Dec 14 09:21:59 2007 From: swissslinky at gmail.com (Dan) Date: Fri, 14 Dec 2007 10:21:59 +0100 Subject: [et-mgmt-tools] Cobbler 7.0.1 FC WebUI Testing on EL4 Message-ID: <1197624119.6683.8.camel@slinky> I skimmed the code and I don't see why it's unexpected? It's being told that it will be coming through...? Note, striking F5 a few times wakes it up. -- Mod_python error: "PythonHandler index" Traceback (most recent call last): File "/usr/lib/python2.3/site-packages/mod_python/apache.py", line 299, in HandlerDispatch result = object(req) File "/var/www/cobbler/web/index.py", line 89, in handler server = "http://127.0.0.1/cobbler_api_rw" TypeError: __init__() got an unexpected keyword argument 'apache' -- Will keep playing :) From markmc at redhat.com Fri Dec 14 12:11:17 2007 From: markmc at redhat.com (Mark McLoughlin) Date: Fri, 14 Dec 2007 12:11:17 +0000 Subject: [et-mgmt-tools] [PATCH 5 of 5] Package/unpackage an image for distribution In-Reply-To: <1197567501.3124.4.camel@localhost.localdomain> References: <1197541713.4428.12.camel@blaa> <1197567501.3124.4.camel@localhost.localdomain> Message-ID: <1197634277.2886.33.camel@blaa> On Thu, 2007-12-13 at 09:38 -0800, David Lutterkort wrote: > On Thu, 2007-12-13 at 10:28 +0000, Mark McLoughlin wrote: > > Interesting ... one thought is that you should expose the Package[1] > > class, and an UnWarePackage sub-class, from virtinst instead of UnWare. > > Also, I'd really only expect a single public Package method - pack(). > > > > Basically, I'm thinking about how the likes of livecd's image-creator > > could use this stuff. > > Yeah, I structured it that way because I didn't think there would be > many users for it - how would you include that with image-creator ? Are > you thinking of doing the image creation soup-to-nuts (kickstart to > tarball, really) in one big jump ? (soup-to-nuts? wtf? :-) But sure, why wouldn't an image creator not dump out the image in a distributable format rather than an intermediate format? I think of the image creator as being similar to rpmbuild in that respect. Also, it's partly about what API we want to add to virtinst and retain compatibility for in future versions. A simple Package base-class, a VMWarePackage sub-class and a pack() method sounds about the extent of the API we'd want to commit to at this point. Cheers, Mark. From mdehaan at redhat.com Fri Dec 14 15:57:12 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 14 Dec 2007 10:57:12 -0500 Subject: [et-mgmt-tools] Cobbler 7.0.1 FC WebUI Testing on EL4 In-Reply-To: <1197624119.6683.8.camel@slinky> References: <1197624119.6683.8.camel@slinky> Message-ID: <4762A7D8.2010709@redhat.com> Dan wrote: > I skimmed the code and I don't see why it's unexpected? It's being told > that it will be coming through...? > > Note, striking F5 a few times wakes it up. > Strange. I'm firing up EL4 in KVM and I'll check it out. Thanks! --Michael > -- > > Mod_python error: "PythonHandler index" > > Traceback (most recent call last): > > File "/usr/lib/python2.3/site-packages/mod_python/apache.py", line 299, in HandlerDispatch > result = object(req) > > File "/var/www/cobbler/web/index.py", line 89, in handler > server = "http://127.0.0.1/cobbler_api_rw" > > TypeError: __init__() got an unexpected keyword argument 'apache' > > -- > > > Will keep playing :) > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From crobinso at redhat.com Fri Dec 14 15:58:06 2007 From: crobinso at redhat.com (Cole Robinson) Date: Fri, 14 Dec 2007 10:58:06 -0500 Subject: [et-mgmt-tools] [PATCH] virt-manager: Add additional info to disk details section In-Reply-To: <20071210195720.GC1973@redhat.com> References: <475D8A2E.9030100@redhat.com> <20071210195720.GC1973@redhat.com> Message-ID: <4762A80E.2030301@redhat.com> Daniel P. Berrange wrote: > On Mon, Dec 10, 2007 at 01:49:18PM -0500, Cole Robinson wrote: >> The attached patch adds a couple fields to the details section for a disk >> device in virt-manager: a 'Driver' field and a 'Permissions' field. The >> driver field is written as "file" or "tap:aio" etc., and the permissions >> field appears as "Readonly" or "Read/Write", with ", Sharable" tacked on >> the end if applicable. > > I've no problem adding the readonly/readwrite/sharable information. > > I don't want the file:/tap:aio/etc stuff shown. Its an underlying implementation > detail over which the admin has no control - we automatically pick the driver > required according to virt type - and as such its not useful to show and people > aren't going to understand what it means. > > Regards, > Dan. Reworked patch attached removing driver field. - Cole -- Cole Robinson crobinso at redhat.com -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: disk-driver-perms-details-patch-2 URL: From berrange at redhat.com Fri Dec 14 16:02:48 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Fri, 14 Dec 2007 16:02:48 +0000 Subject: [et-mgmt-tools] [PATCH] virt-manager: Add additional info to disk details section In-Reply-To: <4762A80E.2030301@redhat.com> References: <475D8A2E.9030100@redhat.com> <20071210195720.GC1973@redhat.com> <4762A80E.2030301@redhat.com> Message-ID: <20071214160248.GH9211@redhat.com> On Fri, Dec 14, 2007 at 10:58:06AM -0500, Cole Robinson wrote: > Daniel P. Berrange wrote: > > On Mon, Dec 10, 2007 at 01:49:18PM -0500, Cole Robinson wrote: > >> The attached patch adds a couple fields to the details section for a disk > >> device in virt-manager: a 'Driver' field and a 'Permissions' field. The > >> driver field is written as "file" or "tap:aio" etc., and the permissions > >> field appears as "Readonly" or "Read/Write", with ", Sharable" tacked on > >> the end if applicable. > > > > I've no problem adding the readonly/readwrite/sharable information. > > > > I don't want the file:/tap:aio/etc stuff shown. Its an underlying implementation > > detail over which the admin has no control - we automatically pick the driver > > required according to virt type - and as such its not useful to show and people > > aren't going to understand what it means. > > > > Regards, > > Dan. > > Reworked patch attached removing driver field. Thanks, 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 mdehaan at redhat.com Fri Dec 14 22:30:42 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 14 Dec 2007 17:30:42 -0500 Subject: [et-mgmt-tools] Cobbler 7.0.1 FC WebUI Testing on EL4 In-Reply-To: <1197624119.6683.8.camel@slinky> References: <1197624119.6683.8.camel@slinky> Message-ID: <47630412.6070209@redhat.com> Dan wrote: > I skimmed the code and I don't see why it's unexpected? It's being told > that it will be coming through...? > > Note, striking F5 a few times wakes it up. > I just tried to replicate this on a fresh EL4.4 install, using Cobbler 0.7.1, and was able to log in the first time in the WebUI. /var/log/cobbler/cobbler.log and /var/log/httpd/error.log also didn't contain any noise. Also, did you restart cobblerd, httpd prior to testing it?... that might have been the problem depending on whether the Apache mod_python config had taken effect or not yet. Note the new URL is http://server/cobbler/web, not the old CGI URL, though the old CGI should have been removed with the previous RPM. Can you provide any other data to help recreate this? --Michael > -- > > Mod_python error: "PythonHandler index" > > Traceback (most recent call last): > > File "/usr/lib/python2.3/site-packages/mod_python/apache.py", line 299, in HandlerDispatch > result = object(req) > > File "/var/www/cobbler/web/index.py", line 89, in handler > server = "http://127.0.0.1/cobbler_api_rw" > > TypeError: __init__() got an unexpected keyword argument 'apache' > > -- > > > Will keep playing :) > > _______________________________________________ > 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 Fri Dec 14 22:50:48 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 14 Dec 2007 17:50:48 -0500 Subject: [et-mgmt-tools] Cobbler 7.0.1 FC WebUI Testing on EL4 In-Reply-To: <47630412.6070209@redhat.com> References: <1197624119.6683.8.camel@slinky> <47630412.6070209@redhat.com> Message-ID: <476308C8.3080209@redhat.com> Michael DeHaan wrote: > Dan wrote: >> I skimmed the code and I don't see why it's unexpected? It's being told >> that it will be coming through...? >> >> Note, striking F5 a few times wakes it up. > > I just tried to replicate this on a fresh EL4.4 install, using Cobbler > 0.7.1, and was able to log in the first time in the WebUI. > /var/log/cobbler/cobbler.log and /var/log/httpd/error.log also didn't > contain any noise. As pointed out on IRC, I mean 0.7.0 (the RPMs on the web site). I've just now bumped git to start 0.7.1. > > Also, did you restart cobblerd, httpd prior to testing it?... that > might have been the problem depending on whether the Apache > mod_python config had taken effect or not yet. Note the new URL is > http://server/cobbler/web, not the old CGI URL, though > the old CGI should have been removed with the previous RPM. > > Can you provide any other data to help recreate this? > > --Michael >> -- >> >> Mod_python error: "PythonHandler index" >> >> Traceback (most recent call last): >> >> File "/usr/lib/python2.3/site-packages/mod_python/apache.py", line >> 299, in HandlerDispatch >> result = object(req) >> >> File "/var/www/cobbler/web/index.py", line 89, in handler >> server = "http://127.0.0.1/cobbler_api_rw" >> >> TypeError: __init__() got an unexpected keyword argument 'apache' >> >> -- >> >> >> Will keep playing :) >> >> _______________________________________________ >> et-mgmt-tools mailing list >> et-mgmt-tools at redhat.com >> https://www.redhat.com/mailman/listinfo/et-mgmt-tools >> > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From rberger at ibd.com Sun Dec 16 02:41:00 2007 From: rberger at ibd.com (Robert J.Berger) Date: Sat, 15 Dec 2007 18:41:00 -0800 Subject: [et-mgmt-tools] Error when running cft manifest command Message-ID: I did a simple session with cft and then tried to create a manifest and got the following error: # cft manifest apache /usr/lib/ruby/site_ruby/1.8/puppet/util/errors.rb:51:in `fail': Could not create local filebucket: Could not configure for running; got 1 failure(s) (Puppet::Error) from /usr/lib/ruby/site_ruby/1.8/puppet/type/pfilebucket.rb:107:in `mkbucket' from /usr/lib/ruby/site_ruby/1.8/puppet/type/pfilebucket.rb:80:in `bucket' from /usr/lib/ruby/site_ruby/1.8/puppet/type/pfile.rb:79:in `default' from /usr/lib/ruby/site_ruby/1.8/puppet/metatype/attributes.rb:635:in `setdefaults' from /usr/lib/ruby/site_ruby/1.8/puppet/metatype/attributes.rb:120:in `eachattr' from /usr/lib/ruby/site_ruby/1.8/puppet/metatype/attributes.rb:115:in `each' from /usr/lib/ruby/site_ruby/1.8/puppet/metatype/attributes.rb:115:in `eachattr' from /usr/lib/ruby/site_ruby/1.8/puppet/metatype/attributes.rb:625:in `setdefaults' ... 11 levels... from /usr/lib/ruby/site_ruby/1.8/cft/puppet/digest.rb:94:in `transportable' from /usr/lib/ruby/site_ruby/1.8/cft/commands.rb:319:in `execute' from /usr/lib/ruby/site_ruby/1.8/cft/commands.rb:148:in `execute' from /usr/sbin/cft:7 Exit 1 I'm running puppet 0.24, Fedora 8, latest cft 0.2.1.1 on x86_64 Any idea what I'm doing wrong or where I can learn how to properly set up cft? I presume I don't have /etc/puppet set up right. But I was trying cft as a way to learn puppet so its a vicious circle! Thanks! ?????????????????????????????? Robert J. Berger - Internet Bandwidth Development, LLC. Voice: 408-838-8896 eFax: +1-408-490-2868 http://www.ibd.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From swissslinky at gmail.com Mon Dec 17 07:57:05 2007 From: swissslinky at gmail.com (Dan) Date: Mon, 17 Dec 2007 08:57:05 +0100 Subject: [et-mgmt-tools] Understand git pull In-Reply-To: <476308C8.3080209@redhat.com> References: <1197624119.6683.8.camel@slinky> <47630412.6070209@redhat.com> <476308C8.3080209@redhat.com> Message-ID: <1197878225.6854.4.camel@slinky> Awesome, thanks Micheal! I've quoted but changed the title, as it's kinda related ;-) I get this problem now and again and being new to git I'm not too sure what's going on. What does the following translate as? -snip- # git pull remote: Generating pack... remote: Done counting 67 objects. Result has 53 objects. remote: Deltifying 53 objects... remote: 100% (53/53) done Unpacking 53 objects... remote: Total 53 (delta 38), reused 0 (delta 0) 100% (53/53) done * refs/remotes/[branch]: fast forward to branch 'devel' of git://fedora/blah.. old..new: ffd71b1..8e1b4b1 Updating ffd71b1..8e1b4b1 cobbler/webui/master.py: needs update fatal: Entry 'cobbler/webui/master.py' not uptodate. Cannot merge. -snip- this was last night. Its happened before, so I've just left it and come back to it later and its usually been fixed. This morning: -snip Updating ffd71b1..8e1b4b1 cobbler/webui/master.py: needs update fatal: Entry 'cobbler/webui/master.py' not uptodate. Cannot merge. -snip- Dan xx On Fri, 2007-12-14 at 17:50 -0500, Michael DeHaan wrote: > Michael DeHaan wrote: > > Dan wrote: > >> I skimmed the code and I don't see why it's unexpected? It's being told > >> that it will be coming through...? > >> > >> Note, striking F5 a few times wakes it up. > > > > I just tried to replicate this on a fresh EL4.4 install, using Cobbler > > 0.7.1, and was able to log in the first time in the WebUI. > > /var/log/cobbler/cobbler.log and /var/log/httpd/error.log also didn't > > contain any noise. > > As pointed out on IRC, I mean 0.7.0 (the RPMs on the web site). I've > just now bumped git to start 0.7.1. > > > > Also, did you restart cobblerd, httpd prior to testing it?... that > > might have been the problem depending on whether the Apache > > mod_python config had taken effect or not yet. Note the new URL is > > http://server/cobbler/web, not the old CGI URL, though > > the old CGI should have been removed with the previous RPM. > > > > Can you provide any other data to help recreate this? > > > > --Michael > >> -- > >> > >> Mod_python error: "PythonHandler index" > >> > >> Traceback (most recent call last): > >> > >> File "/usr/lib/python2.3/site-packages/mod_python/apache.py", line > >> 299, in HandlerDispatch > >> result = object(req) > >> > >> File "/var/www/cobbler/web/index.py", line 89, in handler > >> server = "http://127.0.0.1/cobbler_api_rw" > >> > >> TypeError: __init__() got an unexpected keyword argument 'apache' > >> > >> -- > >> > >> > >> Will keep playing :) > >> > >> _______________________________________________ > >> et-mgmt-tools mailing list > >> et-mgmt-tools at redhat.com > >> https://www.redhat.com/mailman/listinfo/et-mgmt-tools > >> > > > > _______________________________________________ > > et-mgmt-tools mailing list > > et-mgmt-tools at redhat.com > > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From swissslinky at gmail.com Mon Dec 17 08:28:23 2007 From: swissslinky at gmail.com (Dan) Date: Mon, 17 Dec 2007 09:28:23 +0100 Subject: [et-mgmt-tools] Cobbler 7.0.1 FC WebUI Testing on EL4 In-Reply-To: <476308C8.3080209@redhat.com> References: <1197624119.6683.8.camel@slinky> <47630412.6070209@redhat.com> <476308C8.3080209@redhat.com> Message-ID: <1197880103.6854.6.camel@slinky> On Fri, 2007-12-14 at 17:50 -0500, Michael DeHaan wrote: > Michael DeHaan wrote: > > Dan wrote: > >> I skimmed the code and I don't see why it's unexpected? It's being told > >> that it will be coming through...? > >> > >> Note, striking F5 a few times wakes it up. > > > > I just tried to replicate this on a fresh EL4.4 install, using Cobbler > > 0.7.1, and was able to log in the first time in the WebUI. > > /var/log/cobbler/cobbler.log and /var/log/httpd/error.log also didn't > > contain any noise. > > As pointed out on IRC, I mean 0.7.0 (the RPMs on the web site). I've > just now bumped git to start 0.7.1. Merci beaucoup! All fixed. > > > > Also, did you restart cobblerd, httpd prior to testing it?... that > > might have been the problem depending on whether the Apache > > mod_python config had taken effect or not yet. Note the new URL is > > http://server/cobbler/web, not the old CGI URL, though > > the old CGI should have been removed with the previous RPM. > > > > Can you provide any other data to help recreate this? > > > > --Michael > >> -- > >> > >> Mod_python error: "PythonHandler index" > >> > >> Traceback (most recent call last): > >> > >> File "/usr/lib/python2.3/site-packages/mod_python/apache.py", line > >> 299, in HandlerDispatch > >> result = object(req) > >> > >> File "/var/www/cobbler/web/index.py", line 89, in handler > >> server = "http://127.0.0.1/cobbler_api_rw" > >> > >> TypeError: __init__() got an unexpected keyword argument 'apache' > >> > >> -- > >> > >> > >> Will keep playing :) > >> > >> _______________________________________________ > >> et-mgmt-tools mailing list > >> et-mgmt-tools at redhat.com > >> https://www.redhat.com/mailman/listinfo/et-mgmt-tools > >> > > > > _______________________________________________ > > et-mgmt-tools mailing list > > et-mgmt-tools at redhat.com > > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From kanarip at kanarip.com Mon Dec 17 09:21:51 2007 From: kanarip at kanarip.com (Jeroen van Meeuwen) Date: Mon, 17 Dec 2007 10:21:51 +0100 Subject: [et-mgmt-tools] Understand git pull In-Reply-To: <1197878225.6854.4.camel@slinky> References: <1197624119.6683.8.camel@slinky> <47630412.6070209@redhat.com> <476308C8.3080209@redhat.com> <1197878225.6854.4.camel@slinky> Message-ID: <47663FAF.9090004@kanarip.com> Dan wrote: > Awesome, thanks Micheal! > > I've quoted but changed the title, as it's kinda related ;-) > > I get this problem now and again and being new to git I'm not too sure > what's going on. What does the following translate as? > It translates as cobbler/webui/master.py has local changes that are not in the repositories metadata (yet). Try 'git diff' to see the changes, or 'git status' to get an overview, 'git commit' to commit them to the local repitory metadata, or 'git reset --hard HEAD' to reset you to the latest point you have from the upstream repositories (warning, you might/will lose changes). Kind regards, Jeroen van Meeuwen -kanarip From crobinso at redhat.com Mon Dec 17 17:40:33 2007 From: crobinso at redhat.com (Cole Robinson) Date: Mon, 17 Dec 2007 12:40:33 -0500 Subject: [et-mgmt-tools] [PATCH] virtinst: Add 'clock' parameter for hvm guest installs Message-ID: <4766B491.8040600@redhat.com> The attached patch adds a 'clock' key to the full virt guest parameters dictionary in virtinst. The main reasoning for this is to specify 'localtime' instead of 'utc' for a windows guest. If no guest type is specified, the generated xml doesn't contain a clock field and uses libvirt's default, which is the current action. Thanks, Cole -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: virtinst-windows-localtime URL: From crobinso at redhat.com Mon Dec 17 17:42:18 2007 From: crobinso at redhat.com (Cole Robinson) Date: Mon, 17 Dec 2007 12:42:18 -0500 Subject: [et-mgmt-tools] [PATCH] virt-manager: Add additional info to disk details section In-Reply-To: <20071214160248.GH9211@redhat.com> References: <475D8A2E.9030100@redhat.com> <20071210195720.GC1973@redhat.com> <4762A80E.2030301@redhat.com> <20071214160248.GH9211@redhat.com> Message-ID: <4766B4FA.4050800@redhat.com> Daniel P. Berrange wrote: >> Reworked patch attached removing driver field. > > Thanks, looks good. > > Dan. I've applied this: http://hg.et.redhat.com/virt/applications/virt-manager--devel?cs=594dbf22bc89 - Cole From berrange at redhat.com Mon Dec 17 17:47:08 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Mon, 17 Dec 2007 17:47:08 +0000 Subject: [et-mgmt-tools] [PATCH] virtinst: Add 'clock' parameter for hvm guest installs In-Reply-To: <4766B491.8040600@redhat.com> References: <4766B491.8040600@redhat.com> Message-ID: <20071217174708.GE32490@redhat.com> On Mon, Dec 17, 2007 at 12:40:33PM -0500, Cole Robinson wrote: > The attached patch adds a 'clock' key to the full virt guest parameters > dictionary in virtinst. The main reasoning for this is to specify 'localtime' > instead of 'utc' for a windows guest. If no guest type is specified, the > generated xml doesn't contain a clock field and uses libvirt's default, which > is the current action. Looks good. You'll need to also update the tests/*.xml files too though to add in info about the new element, otherwise 'python setup.py test' will fail 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 dan.dengate at cern.ch Tue Dec 18 10:21:50 2007 From: dan.dengate at cern.ch (Dan Dengate) Date: Tue, 18 Dec 2007 11:21:50 +0100 Subject: [et-mgmt-tools] Re: et-mgmt-tools Digest, Vol 16, Issue 20 In-Reply-To: <20071217170021.DDDD873695@hormel.redhat.com> References: <20071217170021.DDDD873695@hormel.redhat.com> Message-ID: <1197973310.6767.8.camel@slinky> On Mon, 2007-12-17 at 12:00 -0500, et-mgmt-tools-request at redhat.com wrote: > Send et-mgmt-tools mailing list submissions to > et-mgmt-tools at redhat.com > > To subscribe or unsubscribe via the World Wide Web, visit > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > or, via email, send a message with subject or body 'help' to > et-mgmt-tools-request at redhat.com > > You can reach the person managing the list at > et-mgmt-tools-owner at redhat.com > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of et-mgmt-tools digest..." > > > Today's Topics: > > 1. Understand git pull (Dan) > 2. Re: Cobbler 7.0.1 FC WebUI Testing on EL4 (Dan) > 3. Re: Understand git pull (Jeroen van Meeuwen) > > > ---------------------------------------------------------------------- > > Message: 1 > Date: Mon, 17 Dec 2007 08:57:05 +0100 > From: Dan > Subject: [et-mgmt-tools] Understand git pull > To: Fedora/Linux Management Tools > Message-ID: <1197878225.6854.4.camel at slinky> > Content-Type: text/plain > > Awesome, thanks Micheal! > > I've quoted but changed the title, as it's kinda related ;-) > > I get this problem now and again and being new to git I'm not too sure > what's going on. What does the following translate as? > > -snip- > # git pull > remote: Generating pack... > remote: Done counting 67 objects. > Result has 53 objects. > remote: Deltifying 53 objects... > remote: 100% (53/53) done > Unpacking 53 objects... > remote: Total 53 (delta 38), reused 0 (delta 0) > 100% (53/53) done > * refs/remotes/[branch]: fast forward to branch 'devel' of > git://fedora/blah.. > old..new: ffd71b1..8e1b4b1 > Updating ffd71b1..8e1b4b1 > cobbler/webui/master.py: needs update > fatal: Entry 'cobbler/webui/master.py' not uptodate. Cannot merge. > > -snip- > > this was last night. Its happened before, so I've just left it and come > back to it later and its usually been fixed. This morning: > > -snip > Updating ffd71b1..8e1b4b1 > cobbler/webui/master.py: needs update > fatal: Entry 'cobbler/webui/master.py' not uptodate. Cannot merge. > -snip- > > > Dan > xx > > On Fri, 2007-12-14 at 17:50 -0500, Michael DeHaan wrote: > > Michael DeHaan wrote: > > > Dan wrote: > > >> I skimmed the code and I don't see why it's unexpected? It's being told > > >> that it will be coming through...? > > >> > > >> Note, striking F5 a few times wakes it up. > > > > > > I just tried to replicate this on a fresh EL4.4 install, using Cobbler > > > 0.7.1, and was able to log in the first time in the WebUI. > > > /var/log/cobbler/cobbler.log and /var/log/httpd/error.log also didn't > > > contain any noise. > > > > As pointed out on IRC, I mean 0.7.0 (the RPMs on the web site). I've > > just now bumped git to start 0.7.1. > > > > > > Also, did you restart cobblerd, httpd prior to testing it?... that > > > might have been the problem depending on whether the Apache > > > mod_python config had taken effect or not yet. Note the new URL is > > > http://server/cobbler/web, not the old CGI URL, though > > > the old CGI should have been removed with the previous RPM. > > > > > > Can you provide any other data to help recreate this? > > > > > > --Michael > > >> -- > > >> > > >> Mod_python error: "PythonHandler index" > > >> > > >> Traceback (most recent call last): > > >> > > >> File "/usr/lib/python2.3/site-packages/mod_python/apache.py", line > > >> 299, in HandlerDispatch > > >> result = object(req) > > >> > > >> File "/var/www/cobbler/web/index.py", line 89, in handler > > >> server = "http://127.0.0.1/cobbler_api_rw" > > >> > > >> TypeError: __init__() got an unexpected keyword argument 'apache' > > >> > > >> -- > > >> > > >> > > >> Will keep playing :) > > >> > > >> _______________________________________________ > > >> et-mgmt-tools mailing list > > >> et-mgmt-tools at redhat.com > > >> https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > >> > > > > > > _______________________________________________ > > > et-mgmt-tools mailing list > > > et-mgmt-tools at redhat.com > > > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > > > _______________________________________________ > > et-mgmt-tools mailing list > > et-mgmt-tools at redhat.com > > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > > > ------------------------------ > > Message: 2 > Date: Mon, 17 Dec 2007 09:28:23 +0100 > From: Dan > Subject: Re: [et-mgmt-tools] Cobbler 7.0.1 FC WebUI Testing on EL4 > To: Fedora/Linux Management Tools > Message-ID: <1197880103.6854.6.camel at slinky> > Content-Type: text/plain > > > On Fri, 2007-12-14 at 17:50 -0500, Michael DeHaan wrote: > > Michael DeHaan wrote: > > > Dan wrote: > > >> I skimmed the code and I don't see why it's unexpected? It's being told > > >> that it will be coming through...? > > >> > > >> Note, striking F5 a few times wakes it up. > > > > > > I just tried to replicate this on a fresh EL4.4 install, using Cobbler > > > 0.7.1, and was able to log in the first time in the WebUI. > > > /var/log/cobbler/cobbler.log and /var/log/httpd/error.log also didn't > > > contain any noise. > > > > As pointed out on IRC, I mean 0.7.0 (the RPMs on the web site). I've > > just now bumped git to start 0.7.1. > > Merci beaucoup! All fixed. > > > > > > > Also, did you restart cobblerd, httpd prior to testing it?... that > > > might have been the problem depending on whether the Apache > > > mod_python config had taken effect or not yet. Note the new URL is > > > http://server/cobbler/web, not the old CGI URL, though > > > the old CGI should have been removed with the previous RPM. > > > > > > Can you provide any other data to help recreate this? > > > > > > --Michael > > >> -- > > >> > > >> Mod_python error: "PythonHandler index" > > >> > > >> Traceback (most recent call last): > > >> > > >> File "/usr/lib/python2.3/site-packages/mod_python/apache.py", line > > >> 299, in HandlerDispatch > > >> result = object(req) > > >> > > >> File "/var/www/cobbler/web/index.py", line 89, in handler > > >> server = "http://127.0.0.1/cobbler_api_rw" > > >> > > >> TypeError: __init__() got an unexpected keyword argument 'apache' > > >> > > >> -- > > >> > > >> > > >> Will keep playing :) > > >> > > >> _______________________________________________ > > >> et-mgmt-tools mailing list > > >> et-mgmt-tools at redhat.com > > >> https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > >> > > > > > > _______________________________________________ > > > et-mgmt-tools mailing list > > > et-mgmt-tools at redhat.com > > > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > > > _______________________________________________ > > et-mgmt-tools mailing list > > et-mgmt-tools at redhat.com > > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > > > ------------------------------ > > Message: 3 > Date: Mon, 17 Dec 2007 10:21:51 +0100 > From: Jeroen van Meeuwen > Subject: Re: [et-mgmt-tools] Understand git pull > To: Fedora/Linux Management Tools > Message-ID: <47663FAF.9090004 at kanarip.com> > Content-Type: text/plain; charset=ISO-8859-1; format=flowed > > Dan wrote: > > Awesome, thanks Micheal! > > > > I've quoted but changed the title, as it's kinda related ;-) > > > > I get this problem now and again and being new to git I'm not too sure > > what's going on. What does the following translate as? > > > > It translates as cobbler/webui/master.py has local changes that are not > in the repositories metadata (yet). > > Try 'git diff' to see the changes, or 'git status' to get an overview, > 'git commit' to commit them to the local repitory metadata, or 'git > reset --hard HEAD' to reset you to the latest point you have from the > upstream repositories (warning, you might/will lose changes). I see, thanks Jeroen!! I now see the error in my ways. I now have my own branch :) Seeing as this could be the last thing I send up before I go on holiday (bring on 5pm!!), to everybody in the Cobbler (and virt) community, Joyeux Noel et Bonne ann?e!! > > Kind regards, > > Jeroen van Meeuwen > -kanarip > > > > > ------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > > End of et-mgmt-tools Digest, Vol 16, Issue 20 > ********************************************* From tim.verhoeven.be at gmail.com Tue Dec 18 14:42:00 2007 From: tim.verhoeven.be at gmail.com (Tim Verhoeven) Date: Tue, 18 Dec 2007 15:42:00 +0100 Subject: [et-mgmt-tools] [PATCH] Cobbler - webui - re-enable random mac js Message-ID: <2a7fce340712180642p71931c48mf14a6d3ff30e1402@mail.gmail.com> Hello, Attached is a small patch that re-enables the "Random" link in the WebUI to generate a random MAC address when clicked. The patch is against a up2date git devel branch. Regards, Tim -- Tim Verhoeven - tim.verhoeven.be at gmail.com - 0479 / 88 11 83 Hoping the problem magically goes away by ignoring it is the "microsoft approach to programming" and should never be allowed. (Linus Torvalds) -------------- next part -------------- A non-text attachment was scrubbed... Name: cobbler-webui-randommac.patch Type: text/x-patch Size: 1540 bytes Desc: not available URL: From crobinso at redhat.com Tue Dec 18 14:57:14 2007 From: crobinso at redhat.com (Cole Robinson) Date: Tue, 18 Dec 2007 09:57:14 -0500 Subject: [et-mgmt-tools] [PATCH] virtinst: Add 'clock' parameter for hvm guest installs In-Reply-To: <20071217174708.GE32490@redhat.com> References: <4766B491.8040600@redhat.com> <20071217174708.GE32490@redhat.com> Message-ID: <4767DFCA.2030008@redhat.com> Daniel P. Berrange wrote: > On Mon, Dec 17, 2007 at 12:40:33PM -0500, Cole Robinson wrote: >> The attached patch adds a 'clock' key to the full virt guest parameters >> dictionary in virtinst. The main reasoning for this is to specify 'localtime' >> instead of 'utc' for a windows guest. If no guest type is specified, the >> generated xml doesn't contain a clock field and uses libvirt's default, which >> is the current action. > > Looks good. You'll need to also update the tests/*.xml files too though to > add in info about the new element, otherwise 'python setup.py test' > will fail > > Regards, > Dan. I committed this yesterday with the appropriate test changes. http://hg.et.redhat.com/virt/applications/virtinst--devel?cs=325a56861208 - Cole From mdehaan at redhat.com Tue Dec 18 15:29:57 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 18 Dec 2007 10:29:57 -0500 Subject: [et-mgmt-tools] [PATCH] Cobbler - webui - re-enable random mac js In-Reply-To: <2a7fce340712180642p71931c48mf14a6d3ff30e1402@mail.gmail.com> References: <2a7fce340712180642p71931c48mf14a6d3ff30e1402@mail.gmail.com> Message-ID: <4767E775.10403@redhat.com> Tim Verhoeven wrote: > Hello, > > Attached is a small patch that re-enables the "Random" link in the > WebUI to generate a random MAC address when clicked. > > The patch is against a up2date git devel branch. > > Regards, > Tim > Multi-interface aware and everything! Nice. Applied! --Michael > > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From agx at sigxcpu.org Tue Dec 18 15:34:14 2007 From: agx at sigxcpu.org (Guido Guenther) Date: Tue, 18 Dec 2007 16:34:14 +0100 Subject: [et-mgmt-tools] virt-manager: handling of bonding and bridge devices Message-ID: <20071218153414.GA16706@bogon.ms20.nix> Hi, currently virt-manager uses some linux specific heuristics to find out if a device is part of a bridge. This immediatley fails if the underlying device is not a physical device but e.g. a bonding device. Hal doesn't seem to care about either bridges nor bonding devices. I wonder where to fix this? Add another linux specific bonding hack to determine if a eth devices is part of a bonding device which might be part of a bridge? Would be easy enough to do but I wonder if it wouldn't be better if hal would export this information in the first place? Cheers, -- Guido From dlutter at redhat.com Tue Dec 18 19:23:14 2007 From: dlutter at redhat.com (David Lutterkort) Date: Tue, 18 Dec 2007 11:23:14 -0800 Subject: [et-mgmt-tools] Error when running cft manifest command In-Reply-To: References: Message-ID: <1198005794.3886.19.camel@localhost.localdomain> On Sat, 2007-12-15 at 18:41 -0800, Robert J.Berger wrote: > > I'm running puppet 0.24, Fedora 8, latest cft 0.2.1.1 on x86_64 > There were actually quite a few things broken in cft for puppet 0.24.0 - I fixed them now and released cft-0.2.2. > Any idea what I'm doing wrong or where I can learn how to properly set > up cft? cft-0.2.1 was definitely broken on puppet 0.24.0; the fixes in cft-0.2.2 seems to make things work properly again, with the stock puppet config that the puppet rpm installs. If you still have trouble after downloading 0.2.2 [1], let me know; I might need a tarball of your /tmp/cft dir then. (Probably better to send privately since it can be quite big) David [1] From the usual places: http://people.redhat.com/dlutter/yum for Fedora RPM's, http://cft.et.redhat.com/downloads for sources From eventi at gridapp.com Tue Dec 18 21:31:35 2007 From: eventi at gridapp.com (Eugene Ventimiglia) Date: Tue, 18 Dec 2007 16:31:35 -0500 Subject: [et-mgmt-tools] cobbler: right place for custom kickstarts In-Reply-To: <4766B491.8040600@redhat.com> References: <4766B491.8040600@redhat.com> Message-ID: We're currently using a homebrew system to provision vms for developers, and I'm in the process of switching that to cobbler. I used to keep a giant monolithic /tftboot/pxelinux.cfg/default with entries for each os we need to install, ie: LABEL manager_plain64 kernel /bootimages/RH-ES-4.0-U4-AMD64/vmlinuz append initrd=/bootimages/RH-ES-4.0-U4-AMD64/initrd.img \ ks=http://osserver/profiles/Linux/manager_plain64.cfg LABEL agent_rh5_vm kernel /bootimages/RH-5.0/vmlinuz append initrd=/bootimages/RH-5.0/initrd.img ks=http://osserver/profiles/Linux/build_rh5_agent.cfg LABEL agent64_rh5_vm kernel /bootimages/RH-5.0-64/vmlinuz append initrd=/bootimages/RH-5.0-64/initrd.img ks=http://osserver/profiles/Linux/build_rh5_agent64.cfg I've played with cobbler for the past day or so, and while I've had great success setting of new machines, I havn't been able to figure out where to put these custom kickstart files. I've read the f manual, and it doesn't say either --e From rigg0022 at umn.edu Tue Dec 18 22:05:17 2007 From: rigg0022 at umn.edu (Riggs, Ben) Date: Tue, 18 Dec 2007 16:05:17 -0600 Subject: [et-mgmt-tools] cobbler: right place for custom kickstarts In-Reply-To: References: <4766B491.8040600@redhat.com> Message-ID: <4768441D.1080806@umn.edu> Generally, the custom kickstarts are in /etc/cobbler/custom/, but they certainly don't have to be. Just specify the path in the profile and you can put them wherever you want. Avoid putting them in /var/www/cobbler/, however, as it is self-cleaning. Ben Eugene Ventimiglia wrote: > We're currently using a homebrew system to provision vms for developers, > and I'm in the process of switching that to cobbler. > > I used to keep a giant monolithic /tftboot/pxelinux.cfg/default with > entries for each os we need to install, ie: > > LABEL manager_plain64 > kernel /bootimages/RH-ES-4.0-U4-AMD64/vmlinuz > append initrd=/bootimages/RH-ES-4.0-U4-AMD64/initrd.img \ > ks=http://osserver/profiles/Linux/manager_plain64.cfg > > LABEL agent_rh5_vm > kernel /bootimages/RH-5.0/vmlinuz > append initrd=/bootimages/RH-5.0/initrd.img > ks=http://osserver/profiles/Linux/build_rh5_agent.cfg > > LABEL agent64_rh5_vm > kernel /bootimages/RH-5.0-64/vmlinuz > append initrd=/bootimages/RH-5.0-64/initrd.img > ks=http://osserver/profiles/Linux/build_rh5_agent64.cfg > > I've played with cobbler for the past day or so, and while I've had > great success setting of new machines, I havn't been able to figure out > where to put these custom kickstart files. > > I've read the f manual, and it doesn't say either > > --e > > _______________________________________________ > 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 Dec 18 22:06:00 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 18 Dec 2007 17:06:00 -0500 Subject: [et-mgmt-tools] cobbler: right place for custom kickstarts In-Reply-To: <4768441D.1080806@umn.edu> References: <4766B491.8040600@redhat.com> <4768441D.1080806@umn.edu> Message-ID: <47684448.9010508@redhat.com> Riggs, Ben wrote: > Generally, the custom kickstarts are in /etc/cobbler/custom/, but they > certainly don't have to be. Just specify the path in the profile and > you can put them wherever you want. Avoid putting them in > /var/www/cobbler/, however, as it is self-cleaning. I've carved out /var/lib/cobbler/kickstarts for you to put templates in, if you'd like to use that. But yes, it doesn't matter... From rainer at ultra-secure.de Tue Dec 18 22:11:55 2007 From: rainer at ultra-secure.de (Rainer Duffner) Date: Tue, 18 Dec 2007 23:11:55 +0100 Subject: [et-mgmt-tools] cobbler: right place for custom kickstarts In-Reply-To: References: <4766B491.8040600@redhat.com> Message-ID: <617621CE-96F3-434F-B302-97B375CC0CE1@ultra-secure.de> Am 18.12.2007 um 22:31 schrieb Eugene Ventimiglia: > We're currently using a homebrew system to provision vms for > developers, > and I'm in the process of switching that to cobbler. > > I used to keep a giant monolithic /tftboot/pxelinux.cfg/default with > entries for each os we need to install, ie: > > LABEL manager_plain64 > kernel /bootimages/RH-ES-4.0-U4-AMD64/vmlinuz > append initrd=/bootimages/RH-ES-4.0-U4-AMD64/initrd.img \ > ks=http://osserver/profiles/Linux/manager_plain64.cfg > > LABEL agent_rh5_vm > kernel /bootimages/RH-5.0/vmlinuz > append initrd=/bootimages/RH-5.0/initrd.img > ks=http://osserver/profiles/Linux/build_rh5_agent.cfg > > LABEL agent64_rh5_vm > kernel /bootimages/RH-5.0-64/vmlinuz > append initrd=/bootimages/RH-5.0-64/initrd.img > ks=http://osserver/profiles/Linux/build_rh5_agent64.cfg > > I've played with cobbler for the past day or so, and while I've had > great success setting of new machines, I havn't been able to figure > out > where to put these custom kickstart files. > > I've read the f manual, and it doesn't say either > In cobbler, you create a profile referencing a kickstart-template (not a complete kickstart-file). The profile should be as generic as possible (mine generally just select a kickstart-template for now). Then, when you create new systems, you choose a profile and actually fill out the template with data - I have a very long "ksmeta" line (BTW: how can I make the ksmeta-inputfieled larger? Could it be made into a textarea?) that does contain all the partitioning and IP- addressing. I couldn't get the "create partition-table with external script"- thing to work, so I use ksmeta-variables and different profiles/ templates) - it works, too, and our requirements are slightly different for each server anyway... So, you have to distill out what is common in your kickstart-files and turn that into one or more kickstart-templates, create different profiles for these kickstart-templates (you have to spell-out the path on the server to in the input-field, there's no nifty "select file" dialogue). Then, when you create systems, you can specify the details that make- up the different systems. I hope this was clear enough (I'm afraid not) - but it's all in the wiki anyway. My suggestion is to do all pre-production testing in either VMware Workstation (or maybe the new Server 2.0 Beta, because of RHEL5- support), for the simple fact that you will reboot often - and that takes minutes on my blades, compared to 1 second with VMware... I probably saved hours this way. cheers, Rainer -- Rainer Duffner CISSP, LPI, MCSE rainer at ultra-secure.de From mdehaan at redhat.com Tue Dec 18 22:23:18 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 18 Dec 2007 17:23:18 -0500 Subject: [et-mgmt-tools] cobbler: right place for custom kickstarts In-Reply-To: <617621CE-96F3-434F-B302-97B375CC0CE1@ultra-secure.de> References: <4766B491.8040600@redhat.com> <617621CE-96F3-434F-B302-97B375CC0CE1@ultra-secure.de> Message-ID: <47684856.5040103@redhat.com> Rainer Duffner wrote: > > Am 18.12.2007 um 22:31 schrieb Eugene Ventimiglia: > >> We're currently using a homebrew system to provision vms for developers, >> and I'm in the process of switching that to cobbler. >> >> I used to keep a giant monolithic /tftboot/pxelinux.cfg/default with >> entries for each os we need to install, ie: >> >> LABEL manager_plain64 >> kernel /bootimages/RH-ES-4.0-U4-AMD64/vmlinuz >> append initrd=/bootimages/RH-ES-4.0-U4-AMD64/initrd.img \ >> ks=http://osserver/profiles/Linux/manager_plain64.cfg >> >> LABEL agent_rh5_vm >> kernel /bootimages/RH-5.0/vmlinuz >> append initrd=/bootimages/RH-5.0/initrd.img >> ks=http://osserver/profiles/Linux/build_rh5_agent.cfg >> >> LABEL agent64_rh5_vm >> kernel /bootimages/RH-5.0-64/vmlinuz >> append initrd=/bootimages/RH-5.0-64/initrd.img >> ks=http://osserver/profiles/Linux/build_rh5_agent64.cfg >> >> I've played with cobbler for the past day or so, and while I've had >> great success setting of new machines, I havn't been able to figure out >> where to put these custom kickstart files. >> >> I've read the f manual, and it doesn't say either >> > > > In cobbler, you create a profile referencing a kickstart-template (not > a complete kickstart-file). It takes complete kickstart files too, but yes. The only thing is if you are feeding it one you already have you'll want to escape any $'s with \$, otherwise the templating engine can get just a little bit confused. > The profile should be as generic as possible (mine generally just > select a kickstart-template for now). Only if you want it to be... Systems and profiles simply map "things" to "what those things do". They can be as minimal or as complex as you want. The simplest profile example is just: cobbler profile add --name=foo --distro=FC-8-i386 --kickstart=/etc/cobbler/kickstart_foo.ks cobbler system add --name=X --mac=AA:BB:CC:DD:EE:FF --profile=foo And of course if you do "cobbler import" you get some starter profiles created for free. > Could it be made into a textarea?) Yes, it could :) > I couldn't get the "create partition-table with external script"-thing > to work, so I use ksmeta-variables and different profiles/templates) - > it works, too, and our requirements are slightly different for each > server anyway... > > So, you have to distill out what is common in your kickstart-files and > turn that into one or more kickstart-templates, create different > profiles for these kickstart-templates (you have to spell-out the path > on the server to in the input-field, there's no nifty "select file" > dialogue). Right, because you're not uploading a local file, you're picking one on the server that already exists. I expect most people to use the command line, but as interest in the Web UI grows, information like this about usability perceptions /is/ helpful. If uploading kickstart files proves useful enough, that's something we can think about adding. > Then, when you create systems, you can specify the details that > make-up the different systems. > > I hope this was clear enough (I'm afraid not) - but it's all in the > wiki anyway. I think the original poster is ok with that and just wanted to know where to store files, but yes, there's lots of online docs on this. > > My suggestion is to do all pre-production testing in either VMware > Workstation (or maybe the new Server 2.0 Beta, because of > RHEL5-support), for the simple fact that you will reboot often - and > that takes minutes on my blades, compared to 1 second with VMware... > I probably saved hours this way. Or Xen.... or KVM .... both of which are free, open-source, and supported in koan :) > > > cheers, > Rainer From rainer at ultra-secure.de Tue Dec 18 22:40:42 2007 From: rainer at ultra-secure.de (Rainer Duffner) Date: Tue, 18 Dec 2007 23:40:42 +0100 Subject: [et-mgmt-tools] cobbler: right place for custom kickstarts In-Reply-To: <47684856.5040103@redhat.com> References: <4766B491.8040600@redhat.com> <617621CE-96F3-434F-B302-97B375CC0CE1@ultra-secure.de> <47684856.5040103@redhat.com> Message-ID: Am 18.12.2007 um 23:23 schrieb Michael DeHaan: > Rainer Duffner wrote: >> >> >> So, you have to distill out what is common in your kickstart-files >> and turn that into one or more kickstart-templates, create >> different profiles for these kickstart-templates (you have to >> spell-out the path on the server to in the input-field, there's no >> nifty "select file" dialogue). > > Right, because you're not uploading a local file, you're picking > one on the server that already exists. > I expect most people to use the command line, but as interest in > the Web UI grows, information like this about usability > perceptions /is/ helpful. I use the web UI, because it's easier to add all the ksmeta-stuff... If I had to provision 100 systems at once, I'd write a script that produces the cobbler commands though. > If uploading kickstart files proves useful enough, that's something > we can think about adding. Not really. Once you know what you have to do it's no problem. Maybe adding a "Note: this is a path to a file already on the server!" next to it will make it a no-brainer... >> >> My suggestion is to do all pre-production testing in either VMware >> Workstation (or maybe the new Server 2.0 Beta, because of RHEL5- >> support), for the simple fact that you will reboot often - and >> that takes minutes on my blades, compared to 1 second with VMware... >> I probably saved hours this way. > > Or Xen.... or KVM .... both of which are free, open-source, and > supported in koan :) > Sorry, I didn't want to discriminate. I just can't get my head around Xen ;-) I will try it one day.... cheers, Rainer -- Rainer Duffner CISSP, LPI, MCSE rainer at ultra-secure.de From agx at sigxcpu.org Wed Dec 19 08:44:12 2007 From: agx at sigxcpu.org (Guido Guenther) Date: Wed, 19 Dec 2007 09:44:12 +0100 Subject: [et-mgmt-tools] [PATCH] virtinst: recommend virt-viewer Message-ID: <20071219084412.GA21990@bogon.ms20.nix> Hi, instead of presenting a python exception recommend installing virt-viewer to the user. Can this be applied? Cheers, -- Guido From agx at sigxcpu.org Mon Dec 17 13:42:01 2007 From: agx at sigxcpu.org (Guido Guenther) Date: Mon, 17 Dec 2007 14:42:01 +0100 Subject: [et-mgmt-tools] [PATCH] don't fail if virt-viewer is missing Message-ID: diff --git a/virt-install b/virt-install index 995f2c3..1215420 100755 --- a/virt-install +++ b/virt-install @@ -15,6 +15,7 @@ import os, sys, string +import errno from optparse import OptionParser, OptionValueError import subprocess import logging @@ -311,7 +312,15 @@ def vnc_console(dom, uri): args = args + [ "--wait", "%s" % dom.ID()] child = os.fork() if not child: - os.execvp(args[0], args) + try: + os.execvp(args[0], args) + except OSError, (err, msg): + if err == errno.ENOENT: + print _("virt-viewer not found, please install the 'virt-viewer' package") + else: + raise OSError, (err, msg) + except e: + raise os._exit(1) return child --UlVJffcvxoiEqYs2-- From crobinso at redhat.com Wed Dec 19 16:20:37 2007 From: crobinso at redhat.com (Cole Robinson) Date: Wed, 19 Dec 2007 11:20:37 -0500 Subject: [et-mgmt-tools] [PATCH] don't fail if virt-viewer is missing In-Reply-To: <20071219155200.5D5F1730D3@hormel.redhat.com> References: <20071219155200.5D5F1730D3@hormel.redhat.com> Message-ID: <476944D5.9030804@redhat.com> Guido Guenther wrote: > diff --git a/virt-install b/virt-install > index 995f2c3..1215420 100755 > --- a/virt-install > +++ b/virt-install > @@ -15,6 +15,7 @@ > > > import os, sys, string > +import errno > from optparse import OptionParser, OptionValueError > import subprocess > import logging > @@ -311,7 +312,15 @@ def vnc_console(dom, uri): > args = args + [ "--wait", "%s" % dom.ID()] > child = os.fork() > if not child: > - os.execvp(args[0], args) > + try: > + os.execvp(args[0], args) > + except OSError, (err, msg): > + if err == errno.ENOENT: > + print _("virt-viewer not found, please install the 'virt-viewer' package") > + else: > + raise OSError, (err, msg) > + except e: > + raise > os._exit(1) > > return child > > --UlVJffcvxoiEqYs2-- > Yes this looks good, I'll apply this. This also brings up another issue that the virtinst rpm requires virt-viewer when we probably shouldn't. virt-viewer indirectly pulls in X (see bz 387971), and virtinst should be perfectly happy running headless. Thanks, Cole From berrange at redhat.com Wed Dec 19 16:27:33 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 19 Dec 2007 16:27:33 +0000 Subject: [et-mgmt-tools] [PATCH] don't fail if virt-viewer is missing In-Reply-To: <476944D5.9030804@redhat.com> References: <20071219155200.5D5F1730D3@hormel.redhat.com> <476944D5.9030804@redhat.com> Message-ID: <20071219162733.GH2973@redhat.com> On Wed, Dec 19, 2007 at 11:20:37AM -0500, Cole Robinson wrote: > Guido Guenther wrote: > > diff --git a/virt-install b/virt-install > > index 995f2c3..1215420 100755 > > --- a/virt-install > > +++ b/virt-install > > @@ -15,6 +15,7 @@ > > > > > > import os, sys, string > > +import errno > > from optparse import OptionParser, OptionValueError > > import subprocess > > import logging > > @@ -311,7 +312,15 @@ def vnc_console(dom, uri): > > args = args + [ "--wait", "%s" % dom.ID()] > > child = os.fork() > > if not child: > > - os.execvp(args[0], args) > > + try: > > + os.execvp(args[0], args) > > + except OSError, (err, msg): > > + if err == errno.ENOENT: > > + print _("virt-viewer not found, please install the 'virt-viewer' package") > > + else: > > + raise OSError, (err, msg) > > + except e: > > + raise > > os._exit(1) > > > > return child > > > > --UlVJffcvxoiEqYs2-- > > > > Yes this looks good, I'll apply this. This also brings up another issue that > the virtinst rpm requires virt-viewer when we probably shouldn't. virt-viewer > indirectly pulls in X (see bz 387971), and virtinst should be perfectly happy > running headless. If you remove virt-viewer as a dep though, then an out of the box install of Fedora won't have virt-viewer and thus all virt-install attempts will fail until they manually install it which sucks. 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 crobinso at redhat.com Wed Dec 19 16:45:13 2007 From: crobinso at redhat.com (Cole Robinson) Date: Wed, 19 Dec 2007 11:45:13 -0500 Subject: [et-mgmt-tools] [PATCH] don't fail if virt-viewer is missing In-Reply-To: <20071219162733.GH2973@redhat.com> References: <20071219155200.5D5F1730D3@hormel.redhat.com> <476944D5.9030804@redhat.com> <20071219162733.GH2973@redhat.com> Message-ID: <47694A99.8000009@redhat.com> Daniel P. Berrange wrote: > On Wed, Dec 19, 2007 at 11:20:37AM -0500, Cole Robinson wrote: >> Guido Guenther wrote: >>> diff --git a/virt-install b/virt-install >>> index 995f2c3..1215420 100755 >>> --- a/virt-install >>> +++ b/virt-install >>> @@ -15,6 +15,7 @@ >>> >>> >>> import os, sys, string >>> +import errno >>> from optparse import OptionParser, OptionValueError >>> import subprocess >>> import logging >>> @@ -311,7 +312,15 @@ def vnc_console(dom, uri): >>> args = args + [ "--wait", "%s" % dom.ID()] >>> child = os.fork() >>> if not child: >>> - os.execvp(args[0], args) >>> + try: >>> + os.execvp(args[0], args) >>> + except OSError, (err, msg): >>> + if err == errno.ENOENT: >>> + print _("virt-viewer not found, please install the 'virt-viewer' package") >>> + else: >>> + raise OSError, (err, msg) >>> + except e: >>> + raise >>> os._exit(1) >>> >>> return child >>> >>> --UlVJffcvxoiEqYs2-- >>> >> Yes this looks good, I'll apply this. This also brings up another issue that >> the virtinst rpm requires virt-viewer when we probably shouldn't. virt-viewer >> indirectly pulls in X (see bz 387971), and virtinst should be perfectly happy >> running headless. > > If you remove virt-viewer as a dep though, then an out of the box install > of Fedora won't have virt-viewer and thus all virt-install attempts will > fail until they manually install it which sucks. > > > Dan. I agree it does suck, though that is the way things were when we tried to exec vncviewer (more common an app I'm sure, but same out of the box result, 'vncviewer not installed'.) I think the inconvenience of having to 'yum install virt-viewer' is less worrying than indirectly requiring X, so that's my vote. - Cole From crobinso at redhat.com Wed Dec 19 17:35:27 2007 From: crobinso at redhat.com (Cole Robinson) Date: Wed, 19 Dec 2007 12:35:27 -0500 Subject: [et-mgmt-tools] [PATCH] don't fail if virt-viewer is missing In-Reply-To: <20071219155200.5D5F1730D3@hormel.redhat.com> References: <20071219155200.5D5F1730D3@hormel.redhat.com> Message-ID: <4769565F.6030504@redhat.com> Guido Guenther wrote: > diff --git a/virt-install b/virt-install > index 995f2c3..1215420 100755 > --- a/virt-install > +++ b/virt-install > @@ -15,6 +15,7 @@ > > > import os, sys, string > +import errno > from optparse import OptionParser, OptionValueError > import subprocess > import logging > @@ -311,7 +312,15 @@ def vnc_console(dom, uri): > args = args + [ "--wait", "%s" % dom.ID()] > child = os.fork() > if not child: > - os.execvp(args[0], args) > + try: > + os.execvp(args[0], args) > + except OSError, (err, msg): > + if err == errno.ENOENT: > + print _("virt-viewer not found, please install the 'virt-viewer' package") > + else: > + raise OSError, (err, msg) > + except e: > + raise > os._exit(1) > > return child > > --UlVJffcvxoiEqYs2-- > I committed this with, with a small change to the error message. Thanks! http://hg.et.redhat.com/virt/applications/virtinst--devel?cs=e5476eeb19e4 - Cole From gene.huft at wachovia.com Wed Dec 19 17:44:30 2007 From: gene.huft at wachovia.com (gene.huft at wachovia.com) Date: Wed, 19 Dec 2007 12:44:30 -0500 Subject: [et-mgmt-tools] Running cobbler, tftp issues Message-ID: Gurus, I've just installed cobbler, I PXE booted a test client, and it successfully gets an IP from my cobbler server running DHCP...however, the client gets a "PXE-E32: TFTP open timeout" error. I assume it cannot find or open the pxelinux.0 file, as described in my /etc/dhcpd.conf. My /etc/xinetd.d/tftp file looks like this: service tftp { socket_type = dgram protocol = udp wait = yes user = root server = /usr/sbin/in.tftpd server_args = -s /tftpboot disable = no per_source = 11 cps = 100 2 flags = IPv4 } and the /tftpboot directory does indeed contain pxelinux.0 : ls -lsart /tftpboot/ total 440 8 drwxr-xr-x 27 root root 4096 Dec 18 09:21 .. 4 -rwxr-xr-x 1 root root 8 Dec 19 04:30 test.bin 4 drwxr-xr-x 4 root root 4096 Dec 19 04:30 . 16 -rwxrwxrwx 1 root root 13148 Dec 19 05:10 pxelinux.0 28 -rwxrwxrwx 1 root root 26496 Dec 19 05:10 menu.c32 372 -rw-r--r-- 1 root root 374146 Dec 19 05:10 elilo-3.6-ia64.efi 4 drwxrwxrwx 5 root root 4096 Dec 19 05:10 images 4 drwxrwxrwx 2 root root 4096 Dec 19 06:01 pxelinux.cfg Any ideas? Thanks in advance. -Gene -------------- next part -------------- An HTML attachment was scrubbed... URL: From gene.huft at wachovia.com Wed Dec 19 18:10:14 2007 From: gene.huft at wachovia.com (gene.huft at wachovia.com) Date: Wed, 19 Dec 2007 13:10:14 -0500 Subject: [et-mgmt-tools] RE: Running cobbler, tftp issues Message-ID: Please disregard....cobbler check revealed ipchains issue. However, now onto the next problem...looks like a new post! Gurus, I've just installed cobbler, I PXE booted a test client, and it successfully gets an IP from my cobbler server running DHCP...however, the client gets a "PXE-E32: TFTP open timeout" error. I assume it cannot find or open the pxelinux.0 file, as described in my /etc/dhcpd.conf. My /etc/xinetd.d/tftp file looks like this: service tftp { socket_type = dgram protocol = udp wait = yes user = root server = /usr/sbin/in.tftpd server_args = -s /tftpboot disable = no per_source = 11 cps = 100 2 flags = IPv4 } and the /tftpboot directory does indeed contain pxelinux.0 : ls -lsart /tftpboot/ total 440 8 drwxr-xr-x 27 root root 4096 Dec 18 09:21 .. 4 -rwxr-xr-x 1 root root 8 Dec 19 04:30 test.bin 4 drwxr-xr-x 4 root root 4096 Dec 19 04:30 . 16 -rwxrwxrwx 1 root root 13148 Dec 19 05:10 pxelinux.0 28 -rwxrwxrwx 1 root root 26496 Dec 19 05:10 menu.c32 372 -rw-r--r-- 1 root root 374146 Dec 19 05:10 elilo-3.6-ia64.efi 4 drwxrwxrwx 5 root root 4096 Dec 19 05:10 images 4 drwxrwxrwx 2 root root 4096 Dec 19 06:01 pxelinux.cfg Any ideas? Thanks in advance. -Gene Gene L. Huft Distributed Systems Engineering ------------------------------------------------ 512-310-4104 512-673-0717 (Cell) Gene.Huft at Wachovia.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From agx at sigxcpu.org Thu Dec 20 15:56:55 2007 From: agx at sigxcpu.org (Guido Guenther) Date: Thu, 20 Dec 2007 16:56:55 +0100 Subject: [et-mgmt-tools] [PATCH] virt-manager: fix errorpath in domain.py Message-ID: <20071220155655.GA13964@bogon.ms20.nix> Hi, trivial patch against version 0.5.2 for fixing the error path, otherwise libvirtError is unknown to the interpreter. Please apply: diff --git a/src/virtManager/domain.py b/src/virtManager/domain.py index 513f857..323a478 100644 --- a/src/virtManager/domain.py +++ b/src/virtManager/domain.py @@ -681,7 +681,7 @@ class vmmDomain(gobject.GObject): try: if self.is_active(): self.vm.attachDevice(xml) - except libvirtError, e: + except libvirt.libvirtError, e: device_exception = str(e) index = vmxml.find("") @@ -703,7 +703,7 @@ class vmmDomain(gobject.GObject): if self.is_active(): try: self.vm.detachDevice(dev_xml) - except libvirtError, e: + except libvirt.libvirtError, e: device_exception = str(e) # then the stored XML Cheers, -- Guido From crobinso at redhat.com Thu Dec 20 16:27:02 2007 From: crobinso at redhat.com (Cole Robinson) Date: Thu, 20 Dec 2007 11:27:02 -0500 Subject: [et-mgmt-tools] [PATCH] virt-manager: fix errorpath in domain.py In-Reply-To: <20071220155655.GA13964@bogon.ms20.nix> References: <20071220155655.GA13964@bogon.ms20.nix> Message-ID: <476A97D6.4010209@redhat.com> Guido Guenther wrote: > Hi, > trivial patch against version 0.5.2 for fixing the error path, otherwise > libvirtError is unknown to the interpreter. Please apply: > > diff --git a/src/virtManager/domain.py b/src/virtManager/domain.py > index 513f857..323a478 100644 > --- a/src/virtManager/domain.py > +++ b/src/virtManager/domain.py > @@ -681,7 +681,7 @@ class vmmDomain(gobject.GObject): > try: > if self.is_active(): > self.vm.attachDevice(xml) > - except libvirtError, e: > + except libvirt.libvirtError, e: > device_exception = str(e) > > index = vmxml.find("") > @@ -703,7 +703,7 @@ class vmmDomain(gobject.GObject): > if self.is_active(): > try: > self.vm.detachDevice(dev_xml) > - except libvirtError, e: > + except libvirt.libvirtError, e: > device_exception = str(e) > > # then the stored XML > > Cheers, > -- Guido > Hi Guido, This was actually already fixed in October: http://hg.et.redhat.com/virt/applications/virt-manager--devel?cs=58c46220f2df We haven't pushed any updates yet so it is not in Fedora. If you want to submit patches it is best to base them on upstream. You can find info on this at: http://virt-manager.et.redhat.com/scmrepo.html Thanks, Cole From agx at sigxcpu.org Thu Dec 20 16:57:38 2007 From: agx at sigxcpu.org (Guido Guenther) Date: Thu, 20 Dec 2007 17:57:38 +0100 Subject: [et-mgmt-tools] [PATCH] virt-manager: fix errorpath in domain.py In-Reply-To: <476A97D6.4010209@redhat.com> References: <20071220155655.GA13964@bogon.ms20.nix> <476A97D6.4010209@redhat.com> Message-ID: <20071220165738.GA14725@bogon.ms20.nix> On Thu, Dec 20, 2007 at 11:27:02AM -0500, Cole Robinson wrote: > This was actually already fixed in October: any chance we see a new upstream release soon? > http://hg.et.redhat.com/virt/applications/virt-manager--devel?cs=58c46220f2df > > We haven't pushed any updates yet so it is not in Fedora. If you want to > submit patches it is best to base them on upstream. You can find info on this at: > > http://virt-manager.et.redhat.com/scmrepo.html I'll check the patches against current hg in the future. Thanks, -- Guido -------------- next part -------------- A non-text attachment was scrubbed... Name: signature.asc Type: application/pgp-signature Size: 189 bytes Desc: Digital signature URL: From tobert at gmail.com Fri Dec 21 21:23:47 2007 From: tobert at gmail.com (Al Tobey) Date: Fri, 21 Dec 2007 13:23:47 -0800 Subject: [et-mgmt-tools] func virt module patch Message-ID: <5ac7acb10712211323q76f28d10o6b0c1ce3db84b33d@mail.gmail.com> The attached patch renames the virt module's info() method to state(), then adds a new info() method that returns all of libvirt's info() data. There's a comment in there about converting some long int's to strings. That was the easy path - a better solution would be nice but I didn't see one on google or the python docs. -Al Tobey -------------- next part -------------- A non-text attachment was scrubbed... Name: 0001-rename-info-to-state-and-add-a-new-info-that-c.patch Type: text/x-patch Size: 2494 bytes Desc: not available URL: From mdehaan at redhat.com Fri Dec 21 21:24:15 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 21 Dec 2007 16:24:15 -0500 Subject: [et-mgmt-tools] func virt module patch In-Reply-To: <5ac7acb10712211323q76f28d10o6b0c1ce3db84b33d@mail.gmail.com> References: <5ac7acb10712211323q76f28d10o6b0c1ce3db84b33d@mail.gmail.com> Message-ID: <476C2EFF.5070409@redhat.com> Al Tobey wrote: > The attached patch renames the virt module's info() method to state(), > then adds a new info() method that returns all of libvirt's info() > data. > > There's a comment in there about converting some long int's to > strings. That was the easy path - a better solution would be nice > but I didn't see one on google or the python docs. > > -Al Tobey > You want this list for Func, BTW... http://www.redhat.com/mailman/listinfo/func-list > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From tdiehl at rogueind.com Sun Dec 23 04:53:55 2007 From: tdiehl at rogueind.com (Tom Diehl) Date: Sat, 22 Dec 2007 23:53:55 -0500 (EST) Subject: [et-mgmt-tools] w2k3 on Centos5-xen Message-ID: Hi, I am trying to use virt-manager to setup a w2k3 domU. I have successfully gotten a centos-5 domU running using cobbler and koan, but windows is not cooperating. The windows installer boots and everything is fine up until the point that the installer reboots the machine. When the machine reboots into the graphical installer I get an error that it can no longer find the media in the cdrom. Of course at that point I have to abort the install because the only options are retry or cancel. I do not know if it is important or not but when windows says it is going to reboot, it actually shutsdown and I have to manually restart it. The config file for the w2k3 domU looks like the following: name = "w2k3" uuid = "24375a2a-30d3-5b4e-1076-bc607f73c35e" maxmem = 500 memory = 500 vcpus = 1 builder = "hvm" kernel = "/usr/lib/xen/boot/hvmloader" boot = "c" pae = 1 acpi = 1 apic = 1 on_poweroff = "destroy" on_reboot = "restart" on_crash = "restart" device_model = "/usr/lib64/xen/bin/qemu-dm" sdl = 0 vnc = 1 vncunused = 1 disk = [ "phy:/dev/GEP0/w2k3,hda,w" ] vif = [ "mac=00:16:3e:04:fd:03,bridge=xenbr0,type=ioemu" ] serial = "pty" (geppetto pts10) # rpm -qa | grep virt libvirt-0.2.3-9.el5 libvirt-0.2.3-9.el5 virt-manager-0.4.0-3.el5 libvirt-python-0.2.3-9.el5 python-virtinst-0.103.0-3.el5.centos (geppetto pts10) # The above file is as created by virt-manager. I have been looking through google for the last few hours but no joy. Is virt-manager devel at a point where this is expected to work? If so, can someone give me an idea what I am missing? Regards, -- Tom Diehl tdiehl at rogueind.com Spamtrap address mtd123 at rogueind.com From chris.sarginson at ukfast.net Mon Dec 24 09:18:05 2007 From: chris.sarginson at ukfast.net (Chris Sarginson) Date: Mon, 24 Dec 2007 09:18:05 +0000 Subject: [et-mgmt-tools] w2k3 on Centos5-xen In-Reply-To: References: Message-ID: <476F794D.6050209@ukfast.net> Hi Tom, The way to resolve this is to run virsh dumpxml $id > file.xml edit file.xml at the relevant section to add in the ISO/cdrom path and then run virsh create file.xml Kind regards Chris Sarginson Technical Support UKFast.Net Ltd (t) 0870 111 8866 (f) 0870 458 4545 "The UK's Best Hosting Provider" ISPA Awards 2007, 2006 and 2005 Dedicated Servers - Managed Hosting - Domain Names- http://www.ukfast.net UKFast.Net Ltd, City Tower, Piccadilly Plaza, Manchester, M1 4BT Registered in England. Number 384 5616 Tom Diehl wrote: > Hi, > > I am trying to use virt-manager to setup a w2k3 domU. I have successfully > gotten a centos-5 domU running using cobbler and koan, but windows is not > cooperating. The windows installer boots and everything is fine up until > the point that the installer reboots the machine. > > When the machine reboots into the graphical installer I get an error that > it can no longer find the media in the cdrom. Of course at that point I > have to abort the install because the only options are retry or cancel. > > I do not know if it is important or not but when windows says it is going > to reboot, it actually shutsdown and I have to manually restart it. > > The config file for the w2k3 domU looks like the following: > > name = "w2k3" > uuid = "24375a2a-30d3-5b4e-1076-bc607f73c35e" > maxmem = 500 > memory = 500 > vcpus = 1 > builder = "hvm" > kernel = "/usr/lib/xen/boot/hvmloader" > boot = "c" > pae = 1 > acpi = 1 > apic = 1 > on_poweroff = "destroy" > on_reboot = "restart" > on_crash = "restart" > device_model = "/usr/lib64/xen/bin/qemu-dm" > sdl = 0 > vnc = 1 > vncunused = 1 > disk = [ "phy:/dev/GEP0/w2k3,hda,w" ] > vif = [ "mac=00:16:3e:04:fd:03,bridge=xenbr0,type=ioemu" ] > serial = "pty" > > (geppetto pts10) # rpm -qa | grep virt > libvirt-0.2.3-9.el5 > libvirt-0.2.3-9.el5 > virt-manager-0.4.0-3.el5 > libvirt-python-0.2.3-9.el5 > python-virtinst-0.103.0-3.el5.centos > (geppetto pts10) # > > The above file is as created by virt-manager. > > I have been looking through google for the last few hours but no joy. > > Is virt-manager devel at a point where this is expected to work? If so, can > someone give me an idea what I am missing? > > Regards, > -- Kind regards Chris Sarginson Technical Support UKFast.Net Ltd (t) 0870 111 8866 (f) 0870 458 4545 "The UK's Best Hosting Provider" ISPA Awards 2007, 2006 and 2005 Dedicated Servers - Managed Hosting - Domain Names- http://www.ukfast.net UKFast.Net Ltd, City Tower, Piccadilly Plaza, Manchester, M1 4BT Registered in England. Number 384 5616 From eventi at gridapp.com Mon Dec 24 18:20:25 2007 From: eventi at gridapp.com (Eugene Ventimiglia) Date: Mon, 24 Dec 2007 13:20:25 -0500 Subject: [et-mgmt-tools] cobbler: import of x86_64 repo creates i386 distro In-Reply-To: References: <4766B491.8040600@redhat.com> Message-ID: I'm running: cobbler import --path=/mnt/os-install/exported/media/Linux/RH-4.0-ES-U4-AMD64/ --name=rhel4 and cobbler list reports: ... distro rhel4-i386 profile rhel4-i386 From fj0873gn at aa.jp.fujitsu.com Tue Dec 25 06:37:58 2007 From: fj0873gn at aa.jp.fujitsu.com (Nobuhiro Itou) Date: Tue, 25 Dec 2007 15:37:58 +0900 Subject: [et-mgmt-tools] [PATCH] BZ#251643: Allow to set the cpu pinning for creating domain with virt-manager. Message-ID: <200712251537.BCE52130.0HOK49G8@aa.jp.fujitsu.com> Hi As well as virt-install, I want to set the cpu pinning at the time of creating domain with virt-manager. And, because virtinst changeset 320 was committed, I add the allocation setting of pCPU to virt-manager. some notes are follows: - This feature works with changeset of virtinst is 320 or later. - The specification about the CPU pin parameter follows that of virt-install. - This function can be used only for xen hypervisor. - When 1 pCPU in xen hypervisor, this function is disabled. Signed-off-by: Nobuhiro Itou Thanks, Nobuhiro Itou. -------------- next part -------------- A non-text attachment was scrubbed... Name: virt-manager_cpupin.patch Type: application/octet-stream Size: 30666 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: virtinst_cpupin.patch Type: application/octet-stream Size: 525 bytes Desc: not available URL: From tdiehl at rogueind.com Wed Dec 26 15:47:07 2007 From: tdiehl at rogueind.com (Tom Diehl) Date: Wed, 26 Dec 2007 10:47:07 -0500 (EST) Subject: [et-mgmt-tools] Re: w2k3 on Centos5-xen In-Reply-To: <476F794D.6050209@ukfast.net> References: <476F794D.6050209@ukfast.net> Message-ID: On Mon, 24 Dec 2007, Chris Sarginson wrote: > Hi Tom, > > The way to resolve this is to run virsh dumpxml $id > file.xml > > edit file.xml at the relevant section to add in the ISO/cdrom path and then > run > > virsh create file.xml Thanks Chris, Once I found the format of that file, it worked like a champ. Is this a bug in virt-manager, I would think virt-manager should add this info when you create a new vm but I am new at all of this virtual stuff so I could be wrong. Regards, -- Tom Diehl tdiehl at rogueind.com Spamtrap address mtd123 at rogueind.com > > Kind regards > > Chris Sarginson > Technical Support > UKFast.Net Ltd > > (t) 0870 111 8866 > (f) 0870 458 4545 > > "The UK's Best Hosting Provider" ISPA Awards 2007, 2006 and 2005 > > Dedicated Servers - Managed Hosting - Domain Names- http://www.ukfast.net > > UKFast.Net Ltd, City Tower, Piccadilly Plaza, Manchester, M1 4BT > Registered in England. Number 384 5616 > > Tom Diehl wrote: >> Hi, >> >> I am trying to use virt-manager to setup a w2k3 domU. I have successfully >> gotten a centos-5 domU running using cobbler and koan, but windows is not >> cooperating. The windows installer boots and everything is fine up until >> the point that the installer reboots the machine. >> >> When the machine reboots into the graphical installer I get an error that >> it can no longer find the media in the cdrom. Of course at that point I >> have to abort the install because the only options are retry or cancel. >> >> I do not know if it is important or not but when windows says it is going >> to reboot, it actually shutsdown and I have to manually restart it. >> >> The config file for the w2k3 domU looks like the following: >> >> name = "w2k3" >> uuid = "24375a2a-30d3-5b4e-1076-bc607f73c35e" >> maxmem = 500 >> memory = 500 >> vcpus = 1 >> builder = "hvm" >> kernel = "/usr/lib/xen/boot/hvmloader" >> boot = "c" >> pae = 1 >> acpi = 1 >> apic = 1 >> on_poweroff = "destroy" >> on_reboot = "restart" >> on_crash = "restart" >> device_model = "/usr/lib64/xen/bin/qemu-dm" >> sdl = 0 >> vnc = 1 >> vncunused = 1 >> disk = [ "phy:/dev/GEP0/w2k3,hda,w" ] >> vif = [ "mac=00:16:3e:04:fd:03,bridge=xenbr0,type=ioemu" ] >> serial = "pty" >> >> (geppetto pts10) # rpm -qa | grep virt >> libvirt-0.2.3-9.el5 >> libvirt-0.2.3-9.el5 >> virt-manager-0.4.0-3.el5 >> libvirt-python-0.2.3-9.el5 >> python-virtinst-0.103.0-3.el5.centos >> (geppetto pts10) # >> >> The above file is as created by virt-manager. >> >> I have been looking through google for the last few hours but no joy. >> >> Is virt-manager devel at a point where this is expected to work? If so, can >> someone give me an idea what I am missing? >> >> Regards, >> > >