From chrispaz at gmail.com Sun Sep 2 17:53:54 2007 From: chrispaz at gmail.com (=?ISO-8859-1?Q?Christian_Pazmi=F1o?=) Date: Sun, 2 Sep 2007 12:53:54 -0500 Subject: [et-mgmt-tools] Build from Installation CD's Message-ID: <7fd7f6230709021053m747ebf90u327414e9757bad3a@mail.gmail.com> Hi, I would like to know if I can use cobbler to build a Network Boot Server using only the installations cd's from RHEL 5, if It is possible how could I do it?? Thanks a lot. -- Att. Christian Pazmi?o http://www.naitcorp.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From chrispaz at gmail.com Sun Sep 2 17:54:04 2007 From: chrispaz at gmail.com (=?ISO-8859-1?Q?Christian_Pazmi=F1o?=) Date: Sun, 2 Sep 2007 12:54:04 -0500 Subject: [et-mgmt-tools] Build from Installation CD's Message-ID: <7fd7f6230709021054l6d5dd234q6d1e353dfddee48e@mail.gmail.com> Hi, I would like to know if I can use cobbler to build a Network -- Att. Christian Pazmi?o http://www.naitcorp.com/ -------------- next part -------------- An HTML attachment was scrubbed... URL: From hhoffman at ip-solutions.net Sun Sep 2 18:16:05 2007 From: hhoffman at ip-solutions.net (Harry Hoffman) Date: Sun, 02 Sep 2007 14:16:05 -0400 Subject: [et-mgmt-tools] Build from Installation CD's In-Reply-To: <7fd7f6230709021053m747ebf90u327414e9757bad3a@mail.gmail.com> References: <7fd7f6230709021053m747ebf90u327414e9757bad3a@mail.gmail.com> Message-ID: <46DAFDE5.3070201@ip-solutions.net> I'll answer both of your emails here... It's not possible to install/use cobbler *only* from the RHEL5 installation CDs. Cobbler (& Koan) are not part of the core distribution so they won't come on the CDs but they are available via yum (RHEL5's primary software update mechanism). Download and install the EPEL repo: http://fedoraproject.org/wiki/EPEL/FAQ#howtouse Then install cobbler and koan yum install cobbler koan That's pretty much all there is to it. Here's a wiki article that comes pretty close: https://hosted.fedoraproject.org/projects/cobbler/wiki/HowToCobblerServer Cheers, Harry Christian Pazmi?o wrote: > Hi, > > I would like to know if I can use cobbler to build a Network Boot Server > using only the installations cd's from RHEL 5, if It is possible how > could I do it?? > > Thanks a lot. > > -- > Att. > Christian Pazmi?o > > http://www.naitcorp.com/ > > > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From jboggs at redhat.com Mon Sep 3 03:25:45 2007 From: jboggs at redhat.com (Joey Boggs) Date: Sun, 02 Sep 2007 23:25:45 -0400 Subject: [et-mgmt-tools] cobbler/virt-factory import errors Message-ID: <46DB7EB9.7010305@redhat.com> I built virt factory from the latest git source on Fedora 7 and have picked up the latest few differences in importing but cant figure this one out. Here are the distros I have setup that were scanned from a RHEL 5 install tree. I created a test profile using vf_gen_profile_stub. In my profile.xml file the distribution is set to RHEL-5-GA which matches on of my distros but it doesnt work when importing. [root at dhcp239-193 rhel5tree]# cobbler distro list RHEL-5-GA RHEL-5-GA-i386 RHEL-5-GA-i686-i386 RHEL-5-GA-xen-i386 2 3 jbtest 4 0.0.1 5 RHEL-5-GA 6 2 7 512 8 512 9 10 11 is_either 12 1 13 14 [root at dhcp239-193 rhel5tree]# vf_import jbtest ***** CONFIG RELOAD ***** populating... scanning through: RHEL-5-GA - looking for: RHEL-5-GA - no such distribution: RHEL-5-GA-xen-i386 - no such distribution: RHEL-5-GA-i386 - no such distribution: RHEL-5-GA-x86_64 - no such distribution: RHEL-5-GA-xen-x86_64 - no compatible distributions have been imported for this profile: jbtest [root at dhcp239-193 rhel5tree]# -------------- next part -------------- An HTML attachment was scrubbed... URL: From lfarkas at bppiac.hu Mon Sep 3 15:06:43 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Mon, 03 Sep 2007 17:06:43 +0200 Subject: [et-mgmt-tools] serial console is gray in virt-manager 0.5 Message-ID: <46DC2303.7020509@bppiac.hu> hi, when i start virt-manager-0.5 with kvm i can't see the console until the kvm guest are start AND i exit and restart virt-manager. and i can't see the guest are booting:-( it was worked with previous versions. -- Levente "Si vis pacem para bellum!" From johan.huysmans at inuits.be Tue Sep 4 14:45:35 2007 From: johan.huysmans at inuits.be (Johan Huysmans) Date: Tue, 4 Sep 2007 16:45:35 +0200 Subject: [et-mgmt-tools] koan --virt: lvm support In-Reply-To: References: Message-ID: Hi All, In March this year Fred posted a request on this mailinglist: http://www.redhat.com/archives/et-mgmt-tools/2007-March/msg00119.html What is the status of this request, is it working in the latest version of koan? I noticed that it is now possible to use a logical volume as a disk, but it is also possible to use logical volumes on dom0 as partitions on domU? Thanks for the information, Johan Huysmans From mdehaan at redhat.com Tue Sep 4 15:01:09 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 04 Sep 2007 11:01:09 -0400 Subject: [et-mgmt-tools] cobbler/virt-factory import errors In-Reply-To: <46DB7EB9.7010305@redhat.com> References: <46DB7EB9.7010305@redhat.com> Message-ID: <46DD7335.8040402@redhat.com> Joey Boggs wrote: > I built virt factory from the latest git source on Fedora 7 and have > picked up the latest few differences in importing but cant figure > this one out. > > Here are the distros I have setup that were scanned from a RHEL 5 > install tree. I created a test profile using vf_gen_profile_stub. In > my profile.xml file the distribution is set to RHEL-5-GA which matches > on of my distros but it doesnt work when importing. virt-factory 0.0.3 was prototyped on FC-6, and 0.0.4 is going towards F-7. Multi-distribution support probably won't be well tested until the next release, though it's pretty good in 0.0.4. My advice would be to attempt to build everything from source (0.0.4) as we are more likely to be able to address problems you find with that -- as well as to use F-7 as your guinea-pig OS. From mdehaan at redhat.com Tue Sep 4 15:08:37 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 04 Sep 2007 11:08:37 -0400 Subject: [et-mgmt-tools] koan --virt: lvm support In-Reply-To: References: Message-ID: <46DD74F5.5000506@redhat.com> Johan Huysmans wrote: > Hi All, > > In March this year Fred posted a request on this mailinglist: > http://www.redhat.com/archives/et-mgmt-tools/2007-March/msg00119.html > Implemented. See the documentation for --virt-path in the koan manpage. > What is the status of this request, is it working in the latest version of koan? > > I noticed that it is now possible to use a logical volume as a disk, > but it is also possible to use logical volumes on dom0 as partitions > on domU? > There are three ways storage can be specified: -- basic disk images on the filesystem (default behavior) -- specifying a logical volume group that has free space in it, where then koan will carve a paritition out of it named after the virtual machine This allows for using a common LVM group for all your virtual machines, which is rather useful. -- specifying a specific partition for koan to use as storage. I hope that answers the question! Recently a few folks have been asking for ways to specify multiple "disks" for their configurations ... and that's something I am going to be looking at. > Thanks for the information, > Johan Huysmans > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From johan.huysmans at inuits.be Tue Sep 4 16:40:15 2007 From: johan.huysmans at inuits.be (Johan Huysmans) Date: Tue, 4 Sep 2007 18:40:15 +0200 Subject: [et-mgmt-tools] koan --virt: lvm support In-Reply-To: <46DD74F5.5000506@redhat.com> References: <46DD74F5.5000506@redhat.com> Message-ID: I noticed the koan parameter --virt-path and tried creating a vhost with following command: koan --server --virt --profile CentOS5-xen-i386 -P VolGroup00 The profile and kickstart files are generated by an import from the CentOs5 DVD. This command creates 1 lv. The domU uses this as a disk (xvda) and creates partitions on this. It is working, but not really the thing i want. I want that koan will create multiple lv's on the dom0. Each lv on dom0 is a partition on domU. This means that the lv is directly formatted in the dom0, and you can mount that logical volume like a normal lv (it IS a normal volume) in the dom0. Is this (allready) possible with koan (or virt-inst)? Greetings, Johan Huysmans On 9/4/07, Michael DeHaan wrote: > Johan Huysmans wrote: > > Hi All, > > > > In March this year Fred posted a request on this mailinglist: > > http://www.redhat.com/archives/et-mgmt-tools/2007-March/msg00119.html > > > > Implemented. > > See the documentation for --virt-path in the koan manpage. > > > > What is the status of this request, is it working in the latest version of koan? > > > > I noticed that it is now possible to use a logical volume as a disk, > > but it is also possible to use logical volumes on dom0 as partitions > > on domU? > > > > There are three ways storage can be specified: > > -- basic disk images on the filesystem (default behavior) > -- specifying a logical volume group that has free space in it, where > then koan will carve a paritition out of it named after the virtual machine > This allows for using a common LVM group for all your virtual machines, > which is rather useful. > -- specifying a specific partition for koan to use as storage. > > I hope that answers the question! > > Recently a few folks have been asking for ways to specify multiple > "disks" for their configurations ... and that's something I am going to > be looking at. > > > Thanks for the information, > > Johan Huysmans > > > > _______________________________________________ > > 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 mdehaan at redhat.com Tue Sep 4 17:14:53 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 04 Sep 2007 13:14:53 -0400 Subject: [et-mgmt-tools] koan --virt: lvm support In-Reply-To: References: <46DD74F5.5000506@redhat.com> Message-ID: <46DD928D.9090803@redhat.com> Johan Huysmans wrote: > I noticed the koan parameter --virt-path and tried creating a vhost > with following command: > koan --server --virt --profile CentOS5-xen-i386 -P VolGroup00 > > The profile and kickstart files are generated by an import from the CentOs5 DVD. > > This command creates 1 lv. The domU uses this as a disk (xvda) and > creates partitions on this. > It is working, but not really the thing i want. > > I want that koan will create multiple lv's on the dom0. Each lv on > dom0 is a partition on domU. This means that the lv is directly > formatted in the dom0, and you can mount that logical volume like a > normal lv (it IS a normal volume) in the dom0. > > Is this (allready) possible with koan (or virt-inst)? > I believe you are stuck with the extra level of indirection -- that is that partitions in your host (dom0) appear as disks in your domU -- with both tools. I don't think you can feed a partition up without a corresponding virtual disk :) (Note that you don't have to use LVM for this -- regular partitions work too) --Michael > Greetings, > Johan Huysmans > > > On 9/4/07, Michael DeHaan wrote: > >> Johan Huysmans wrote: >> >>> Hi All, >>> >>> In March this year Fred posted a request on this mailinglist: >>> http://www.redhat.com/archives/et-mgmt-tools/2007-March/msg00119.html >>> >>> >> Implemented. >> >> See the documentation for --virt-path in the koan manpage. >> >> >> >>> What is the status of this request, is it working in the latest version of koan? >>> >>> I noticed that it is now possible to use a logical volume as a disk, >>> but it is also possible to use logical volumes on dom0 as partitions >>> on domU? >>> >>> >> There are three ways storage can be specified: >> >> -- basic disk images on the filesystem (default behavior) >> -- specifying a logical volume group that has free space in it, where >> then koan will carve a paritition out of it named after the virtual machine >> This allows for using a common LVM group for all your virtual machines, >> which is rather useful. >> -- specifying a specific partition for koan to use as storage. >> >> I hope that answers the question! >> >> Recently a few folks have been asking for ways to specify multiple >> "disks" for their configurations ... and that's something I am going to >> be looking at. >> >> >>> Thanks for the information, >>> Johan Huysmans >>> >>> _______________________________________________ >>> 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 johan.huysmans at inuits.be Tue Sep 4 17:46:50 2007 From: johan.huysmans at inuits.be (Johan Huysmans) Date: Tue, 4 Sep 2007 19:46:50 +0200 Subject: [et-mgmt-tools] koan --virt: lvm support In-Reply-To: <46DD928D.9090803@redhat.com> References: <46DD74F5.5000506@redhat.com> <46DD928D.9090803@redhat.com> Message-ID: well... it is possible ;) because we are allready doing this for over a year. And the installation is quite simple: * you create your logical volumes in the dom0 and mount them * you install a minimal system with yum --installroot= * you create a xen config where you map the logical volumes to the partition inside the domU. The installation instructions can be found on my blog: http://www.raskas.be/blog/howtos/centos-base-image/ And this is how the disk section of my xen config looks like: disk = ['phy:vm_volumes/root-host1,sda1,w' ,'phy:vm_volumes/swap-host1,sda2,w' ,'phy:vm_volumes/tmp-host1,sda3,w' ,'phy:vm_volumes/varlog-host1,sda6,w' ,'phy:vm_volumes/home-host1,sda7,w' ] It would be really great if this would be possible to do with koan, for now we do this with our own create vhost script. Greetings, Johan On 9/4/07, Michael DeHaan wrote: > Johan Huysmans wrote: > > I noticed the koan parameter --virt-path and tried creating a vhost > > with following command: > > koan --server --virt --profile CentOS5-xen-i386 -P VolGroup00 > > > > The profile and kickstart files are generated by an import from the CentOs5 DVD. > > > > This command creates 1 lv. The domU uses this as a disk (xvda) and > > creates partitions on this. > > It is working, but not really the thing i want. > > > > I want that koan will create multiple lv's on the dom0. Each lv on > > dom0 is a partition on domU. This means that the lv is directly > > formatted in the dom0, and you can mount that logical volume like a > > normal lv (it IS a normal volume) in the dom0. > > > > Is this (allready) possible with koan (or virt-inst)? > > > > I believe you are stuck with the extra level of indirection -- that is > that partitions in your host (dom0) appear > as disks in your domU -- with both tools. I don't think you can feed a > partition up without a corresponding virtual disk :) > > (Note that you don't have to use LVM for this -- regular partitions work > too) > > --Michael > > > > > > > > > Greetings, > > Johan Huysmans > > > > > > On 9/4/07, Michael DeHaan wrote: > > > >> Johan Huysmans wrote: > >> > >>> Hi All, > >>> > >>> In March this year Fred posted a request on this mailinglist: > >>> http://www.redhat.com/archives/et-mgmt-tools/2007-March/msg00119.html > >>> > >>> > >> Implemented. > >> > >> See the documentation for --virt-path in the koan manpage. > >> > >> > >> > >>> What is the status of this request, is it working in the latest version of koan? > >>> > >>> I noticed that it is now possible to use a logical volume as a disk, > >>> but it is also possible to use logical volumes on dom0 as partitions > >>> on domU? > >>> > >>> > >> There are three ways storage can be specified: > >> > >> -- basic disk images on the filesystem (default behavior) > >> -- specifying a logical volume group that has free space in it, where > >> then koan will carve a paritition out of it named after the virtual machine > >> This allows for using a common LVM group for all your virtual machines, > >> which is rather useful. > >> -- specifying a specific partition for koan to use as storage. > >> > >> I hope that answers the question! > >> > >> Recently a few folks have been asking for ways to specify multiple > >> "disks" for their configurations ... and that's something I am going to > >> be looking at. > >> > >> > >>> Thanks for the information, > >>> Johan Huysmans > >>> > >>> _______________________________________________ > >>> 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 > > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From wright at imageworks.com Tue Sep 4 17:50:34 2007 From: wright at imageworks.com (Peter Wright) Date: Tue, 04 Sep 2007 10:50:34 -0700 Subject: [et-mgmt-tools] Alternative storage backends for Cobbler / otherfeatures In-Reply-To: <46D88E85.3050904@redhat.com> References: <46D88E85.3050904@redhat.com> Message-ID: <46DD9AEA.90400@imageworks.com> Michael DeHaan wrote: > > So there's been a lot of interest around having Cobbler read LDAP > recently, and possibly I'm guessing it would be useful to have it > /write/ to LDAP. I know several > folks already have their own scripts to interface between the two. > This would be a *huge* win for my environment, and I'm sure for other folks dealing with more than a handful of machines. As far as writing to LDAP, while it may be nice in some circumstances - generating config files alone would be excellent. > Most likely what would happen is I'll implement the framework for > allowing arbitrary formats with a sqlite prototype, and if someone else > wants to add > in LDAP later that would be pretty easy to do by following the sqlite > module's lead. > SQL support would be interesting - assuming that most people store their machine inventory in some soft of RDBMS - but for my use I think having LDAP support will scale better to multiple facilities. Enterprise LDAP implementations are generally designed with replication in mind (be it for a local LDAP cluster or to replicate a LDAP domain across multiple sites) - replicating a SQL DB seems like excessive baggage, especially if you are hoping to use cobbler/koan to manage multiple sites and datacenters. > > The other thing on the radar is finally making the XMLRPC API > bi-directional (by adding an additional secure version on another port) > to make the life of webapps using the Cobbler API easier. I've been > meaning to do that for a while. Until then apps that need write access > to cobbler configs can go through the python API and/or the YAML tree. > nice!! > > Comments? Questions? Ideas? > thanks for all the hard work you've done with cobbler/koan - it's literally shaved several weeks of engineering time for a project I am working on! I wish I had more time to help out getting these features added to cobbler, unfortunately that is unlikely for the foreseeable future. I can offer testing, debugging help and as usual my two bit's :) -pete -- Peter Wright Systems Administrator Sony Pictures Imageworks wright at imageworks.com www.imageworks.com From mdehaan at redhat.com Tue Sep 4 17:55:53 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 04 Sep 2007 13:55:53 -0400 Subject: [et-mgmt-tools] koan --virt: lvm support In-Reply-To: References: <46DD74F5.5000506@redhat.com> <46DD928D.9090803@redhat.com> Message-ID: <46DD9C29.5050308@redhat.com> Johan Huysmans wrote: > well... it is possible ;) > because we are allready doing this for over a year. > The word on the street (from Dan Berrange) is that this feature is not guaranteed to work in future versions of Xen. --Michael > And the installation is quite simple: > * you create your logical volumes in the dom0 and mount them > * you install a minimal system with yum --installroot= > * you create a xen config where you map the logical volumes to the > partition inside the domU. > > The installation instructions can be found on my blog: > http://www.raskas.be/blog/howtos/centos-base-image/ > > And this is how the disk section of my xen config looks like: > disk = ['phy:vm_volumes/root-host1,sda1,w' > ,'phy:vm_volumes/swap-host1,sda2,w' > ,'phy:vm_volumes/tmp-host1,sda3,w' > ,'phy:vm_volumes/varlog-host1,sda6,w' > ,'phy:vm_volumes/home-host1,sda7,w' > ] > > It would be really great if this would be possible to do with koan, > for now we do this with our own create vhost script. > > Greetings, > Johan > > On 9/4/07, Michael DeHaan wrote: > >> Johan Huysmans wrote: >> >>> I noticed the koan parameter --virt-path and tried creating a vhost >>> with following command: >>> koan --server --virt --profile CentOS5-xen-i386 -P VolGroup00 >>> >>> The profile and kickstart files are generated by an import from the CentOs5 DVD. >>> >>> This command creates 1 lv. The domU uses this as a disk (xvda) and >>> creates partitions on this. >>> It is working, but not really the thing i want. >>> >>> I want that koan will create multiple lv's on the dom0. Each lv on >>> dom0 is a partition on domU. This means that the lv is directly >>> formatted in the dom0, and you can mount that logical volume like a >>> normal lv (it IS a normal volume) in the dom0. >>> >>> Is this (allready) possible with koan (or virt-inst)? >>> >>> >> I believe you are stuck with the extra level of indirection -- that is >> that partitions in your host (dom0) appear >> as disks in your domU -- with both tools. I don't think you can feed a >> partition up without a corresponding virtual disk :) >> >> (Note that you don't have to use LVM for this -- regular partitions work >> too) >> >> --Michael >> >> >> >> >> >> >> >> >>> Greetings, >>> Johan Huysmans >>> >>> >>> On 9/4/07, Michael DeHaan wrote: >>> >>> >>>> Johan Huysmans wrote: >>>> >>>> >>>>> Hi All, >>>>> >>>>> In March this year Fred posted a request on this mailinglist: >>>>> http://www.redhat.com/archives/et-mgmt-tools/2007-March/msg00119.html >>>>> >>>>> >>>>> >>>> Implemented. >>>> >>>> See the documentation for --virt-path in the koan manpage. >>>> >>>> >>>> >>>> >>>>> What is the status of this request, is it working in the latest version of koan? >>>>> >>>>> I noticed that it is now possible to use a logical volume as a disk, >>>>> but it is also possible to use logical volumes on dom0 as partitions >>>>> on domU? >>>>> >>>>> >>>>> >>>> There are three ways storage can be specified: >>>> >>>> -- basic disk images on the filesystem (default behavior) >>>> -- specifying a logical volume group that has free space in it, where >>>> then koan will carve a paritition out of it named after the virtual machine >>>> This allows for using a common LVM group for all your virtual machines, >>>> which is rather useful. >>>> -- specifying a specific partition for koan to use as storage. >>>> >>>> I hope that answers the question! >>>> >>>> Recently a few folks have been asking for ways to specify multiple >>>> "disks" for their configurations ... and that's something I am going to >>>> be looking at. >>>> >>>> >>>> >>>>> Thanks for the information, >>>>> Johan Huysmans >>>>> >>>>> _______________________________________________ >>>>> 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 >>> >>> >> _______________________________________________ >> 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 mdehaan at redhat.com Tue Sep 4 22:33:07 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 04 Sep 2007 18:33:07 -0400 Subject: [et-mgmt-tools] new feature: creating multiple virtual disks for a koan VM Message-ID: <46DDDD23.3070201@redhat.com> Using the latest source of Cobbler and Koan (which will be 0.6.2), the following examples of adding more than one disk to a virtual machine are now possible. # request three seperate disk images, 5 GB each cobbler profile add --name=foo --distro=bar --kickstart=/etc/cobbler/baz.ks --virt-path=/opt/foo/A,/opt/foo/B,/opt/foo/C --virt-size=5 # request one disk as a loopback in /opt/foo, another on partition /dev/sda4, using 5 GB each cobbler profile add --name=foo --distro=bar --kickstart=/etc/cobbler/baz.ks --virt-path=/opt/foo,/dev/sda4 --virt-size=5 # (the following example actually needs some tweaking but will work soon) # request two disks on VolGroup00 using 5 GB and 6 GB cobbler profile add --name=foo --distro=bar --kickstart=/etc/cobbler/baz.ks --virt-path=VolGroup00,VolGroup00 --virt-size=5,6 As in previous versions, if you don't pick a value for --virt-path in either koan or cobbler, koan will pick a file in either /var/lib/xen/images or /opt/qemu that is named after the MAC address of the virtual system. When you don't set a --virt-path, you get one disk every time. As an aside: here's how you can specify --virt-path in koan to override what was set in cobbler: koan --virt --profile=F-7-i386 --server=bootserver.example.org --virt-path=VolGroup01,VolGroup02 As you can see, the syntax looks exactly like it does in Cobbler. Any value Cobbler had for --virt-path would be overriden. Note that --virt-size still needs to come from Cobbler. Rather intentionally, the number of things that can be overriden in koan is kept to a bare minimum. Note that if you specify only one value for --virt-path in Cobbler (say "5", for "5 Gigabytes") and the user requests two disks, they'll both be 5 gigabytes in size. This is to be able to If you don't need multiple disks, the old commands still work when single values are left ... and the defaults still work the way they used to. If the above doesn't make sense, let me know. In general this is not something most folks will need, but for the ones that do, there you go... Thanks! --Michael From berrange at redhat.com Wed Sep 5 03:28:41 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 5 Sep 2007 04:28:41 +0100 Subject: [et-mgmt-tools] koan --virt: lvm support In-Reply-To: <46DD9C29.5050308@redhat.com> References: <46DD74F5.5000506@redhat.com> <46DD928D.9090803@redhat.com> <46DD9C29.5050308@redhat.com> Message-ID: <20070905032841.GB8649@redhat.com> On Tue, Sep 04, 2007 at 01:55:53PM -0400, Michael DeHaan wrote: > Johan Huysmans wrote: > >well... it is possible ;) > >because we are allready doing this for over a year. > > > > The word on the street (from Dan Berrange) is that this feature is not > guaranteed to work in future > versions of Xen. To be a little more verbose.... Virtual machines are assigned virtual *disks*. ie hdN, sdN (for full virt) or xvdN (for paravirt). Names like sda1 are partitions, and not disks. It makes no conceptual sense to have a partition without an enclosing disk. If you want to have a 1-1 mapping between the backing files & disks inside the guest, then simply don't partition the disk inside the guest - format /dev/sda directly. Xen paravirt only happens to allow you to map a file straight through as a partition, and while it may currently work, I can make absolutely no guarentees it will work in the future as the Xen paravirt kernel evolves - particularly as Xen increasingly merges into LKML trees. NB with the LKML merge the ability to hijack hdN and sdN in paravirt guests has been removed - paravirt can only use xvdN devices now. I've not checked if hijacking partitions as described below works, but I would not be surprised if it doesn't. Finally its also not portable to Xen fullyvirt, or KVM, or QEMU, etc. Basically unsupportable in the medium-to-long term. > >And the installation is quite simple: > >* you create your logical volumes in the dom0 and mount them > >* you install a minimal system with yum --installroot= > >* you create a xen config where you map the logical volumes to the > >partition inside the domU. > > > >The installation instructions can be found on my blog: > >http://www.raskas.be/blog/howtos/centos-base-image/ > > > >And this is how the disk section of my xen config looks like: > >disk = ['phy:vm_volumes/root-host1,sda1,w' > > ,'phy:vm_volumes/swap-host1,sda2,w' > > ,'phy:vm_volumes/tmp-host1,sda3,w' > > ,'phy:vm_volumes/varlog-host1,sda6,w' > > ,'phy:vm_volumes/home-host1,sda7,w' > > ] > > > >It would be really great if this would be possible to do with koan, > >for now we do this with our own create vhost script. > > 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 Sep 5 03:31:14 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 5 Sep 2007 04:31:14 +0100 Subject: [et-mgmt-tools] serial console is gray in virt-manager 0.5 In-Reply-To: <46DC2303.7020509@bppiac.hu> References: <46DC2303.7020509@bppiac.hu> Message-ID: <20070905033114.GC8649@redhat.com> On Mon, Sep 03, 2007 at 05:06:43PM +0200, Farkas Levente wrote: > hi, > when i start virt-manager-0.5 with kvm i can't see the console until the > kvm guest are start AND i exit and restart virt-manager. and i can't see > the guest are booting:-( it was worked with previous versions. When you say the serial console is gray, could you clarify whether you mean the graphical console (aka the embedded VNC widget, which may be gray if not available), or the actual seriual console (which is black, not gray). There are a couple of known issues with the graphical console in 0.5.0 which I have fixed upstream, thanks to a patch from Bernhard Kaindl at Novell. In particular this changeset: http://hg.et.redhat.com/virt/applications/virt-manager--devel?cs=6e055c5f5e19 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 Sep 5 03:33:51 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 5 Sep 2007 04:33:51 +0100 Subject: [et-mgmt-tools] Can't clone a DomU In-Reply-To: References: Message-ID: <20070905033351.GD8649@redhat.com> On Tue, Aug 28, 2007 at 05:09:45PM +0200, Krammer Oliver wrote: > Hi Everybody, > > I hobe i am at the right list. I used virt-clone for cloning a domu i > started the clone process with: > > ./virt-clone -o defaultsuse10 -n suseclone01 -f > /var/lib/xen/images/suseclone01/disk0 -d > This worked after that i got > servxen01:/etc/xen/vm # xm list > Name ID Mem VCPUs State > Time(s) > Domain-0 0 15093 8 r----- 4509.4 > defaultsuse10 13 1024 2 -b---- 24.6 > suseclone01 1024 1 0.0 > > When i start the dom with > xm start suseclone01 > Xen will hang up and dosn't respond any more > > The log file give me following statemantes: > > [2007-08-28 19:08:35 xend.XendDomainInfo 31046] DEBUG (XendDomainInfo:1335) > XendDomainInfo.constructDomain > [2007-08-28 19:08:35 xend 31046] DEBUG (balloon:146) Balloon: 16724 KiB > free; need 16384; done. > [2007-08-28 19:08:35 xend.XendDomainInfo 31046] DEBUG (XendDomainInfo:1382) > XendDomainInfo.initDomain: 17 256 > [2007-08-28 19:08:35 xend 31113] DEBUG (XendBootloader:69) Launching > bootloader as ['/usr/lib/xen/boot/domUloader.py', '-q', '--args=TERM=xterm > ', '--output=/var/run/xend/boot/xenbl.23654', > '/var/lib/xen/images/suseclone01/disk0']. If that's the last line in the xend.log, I'd say the bootloader is hanging while trying to extract the kernel/initrd. What happens if you run the domUloader.py script directly with the args shown in the log ? It sounds like the virt-clone command itself completed successfully but it may be worth checking in /root/.virtinst/virt-clone.log to see if any problems were logged. Alternatively run it in --debug flag which makes it more verbose. 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 fj0588di at aa.jp.fujitsu.com Wed Sep 5 08:15:01 2007 From: fj0588di at aa.jp.fujitsu.com (S.Sakamoto) Date: Wed, 5 Sep 2007 17:15:01 +0900 Subject: [et-mgmt-tools] Can't clone a DomU In-Reply-To: References: Message-ID: <200709051715.ADE13505.09E9JK6G@aa.jp.fujitsu.com> On Tue, Aug 28, 2007 at 05:09:45PM +0200, Krammer Oliver wrote: > Hi Everybody, > > I hobe i am at the right list. I used virt-clone for cloning a domu i > started the clone process with: > > ./virt-clone -o defaultsuse10 -n suseclone01 -f > /var/lib/xen/images/suseclone01/disk0 -d > This worked after that i got > servxen01:/etc/xen/vm # xm list > Name ID Mem VCPUs State > Time(s) > Domain-0 0 15093 8 r----- 4509.4 > defaultsuse10 13 1024 2 -b---- 24.6 > suseclone01 1024 1 0.0 > > When i start the dom with > xm start suseclone01 > Xen will hang up and dosn't respond any more > > The log file give me following statemantes: > > [2007-08-28 19:08:35 xend.XendDomainInfo 31046] DEBUG (XendDomainInfo:1335) > XendDomainInfo.constructDomain > [2007-08-28 19:08:35 xend 31046] DEBUG (balloon:146) Balloon: 16724 KiB > free; need 16384; done. > [2007-08-28 19:08:35 xend.XendDomainInfo 31046] DEBUG (XendDomainInfo:1382) > XendDomainInfo.initDomain: 17 256 > [2007-08-28 19:08:35 xend 31113] DEBUG (XendBootloader:69) Launching > bootloader as ['/usr/lib/xen/boot/domUloader.py', '-q', '--args=TERM=xterm > ', '--output=/var/run/xend/boot/xenbl.23654', > '/var/lib/xen/images/suseclone01/disk0']. > It seems that it isn't a problem of virt-clone as a result that I investigated this problem. So, I want more details of this, take the following information. 1. Does defaultsuse10 which is the original start normaly? 2. The definition file of defaultsuse10(original) and suseclone01(clone). (The definition file is output in "xm list --long" and "virsh dumpxml".) 3. The output when you carried it out with "--debug" or /root/.virtinst/virt-clone.log virt-clone does not modify tag of original. I am just thinking that it is the problem of the boot loader, domUloader.py, of SUSE. Shigeki Sakamoto. -- _/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/ Shigeki Sakamoto Toyama Fujitsu Limited Software Development)2nd E-mail:fj0588di at aa.jp.fujitsu.com _/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/ From mlkb at x-tend.be Wed Sep 5 09:18:49 2007 From: mlkb at x-tend.be (Kris Buytaert) Date: Wed, 05 Sep 2007 11:18:49 +0200 Subject: [et-mgmt-tools] koan --virt: lvm support In-Reply-To: <20070905032841.GB8649@redhat.com> References: <46DD74F5.5000506@redhat.com> <46DD928D.9090803@redhat.com> <46DD9C29.5050308@redhat.com> <20070905032841.GB8649@redhat.com> Message-ID: <1188983929.3750.7.camel@mine3.x-tend.be> On Wed, 2007-09-05 at 04:28 +0100, Daniel P. Berrange wrote: > On Tue, Sep 04, 2007 at 01:55:53PM -0400, Michael DeHaan wrote: > > Johan Huysmans wrote: > > >well... it is possible ;) > > >because we are allready doing this for over a year. > > > > > > > The word on the street (from Dan Berrange) is that this feature is not > > guaranteed to work in future > > versions of Xen. > > To be a little more verbose.... > > Virtual machines are assigned virtual *disks*. ie hdN, sdN (for full virt) or > xvdN (for paravirt). Names like sda1 are partitions, and not disks. It makes > no conceptual sense to have a partition without an enclosing disk. If you want > to have a 1-1 mapping between the backing files & disks inside the guest, then > simply don't partition the disk inside the guest - format /dev/sda directly. > Xen paravirt only happens to allow you to map a file straight through as a > partition, and while it may currently work, I can make absolutely no guarentees > it will work in the future as the Xen paravirt kernel evolves - particularly > as Xen increasingly merges into LKML trees. NB with the LKML merge the ability > to hijack hdN and sdN in paravirt guests has been removed - paravirt can only > use xvdN devices now. I've not checked if hijacking partitions as described > below works, but I would not be surprised if it doesn't. Finally its also not > portable to Xen fullyvirt, or KVM, or QEMU, etc. Basically unsupportable in > the medium-to-long term. I think it is extremely important to continue to be able to use filesystems on LVM volumes in dom0 as filesystems in domU's. Using them as full disks with a partition on them would be fine. greetz -- Kris Buytaert From lfarkas at bppiac.hu Wed Sep 5 10:35:31 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Wed, 05 Sep 2007 12:35:31 +0200 Subject: [et-mgmt-tools] serial console is gray in virt-manager 0.5 In-Reply-To: <20070905033114.GC8649@redhat.com> References: <46DC2303.7020509@bppiac.hu> <20070905033114.GC8649@redhat.com> Message-ID: <46DE8673.3030401@bppiac.hu> Daniel P. Berrange wrote: > On Mon, Sep 03, 2007 at 05:06:43PM +0200, Farkas Levente wrote: >> hi, >> when i start virt-manager-0.5 with kvm i can't see the console until the >> kvm guest are start AND i exit and restart virt-manager. and i can't see >> the guest are booting:-( it was worked with previous versions. > > When you say the serial console is gray, could you clarify whether you mean > the graphical console (aka the embedded VNC widget, which may be gray if > not available), or the actual seriual console (which is black, not gray). the graphical console (ie the embeded vnc widget is working, but just after a reboot) but the 3th manu item "serial consol" is gray. is there any way to get it working? or if it's never working why there is such a menu item? -- Levente "Si vis pacem para bellum!" From Kevin.Maute at WPAFB.AF.MIL Wed Sep 5 14:34:24 2007 From: Kevin.Maute at WPAFB.AF.MIL (Maute, Kevin P CTR AFRL/VAOO) Date: Wed, 5 Sep 2007 10:34:24 -0400 Subject: [et-mgmt-tools] FW: Help with cobbler for dummies... Message-ID: <29A7C5F65D54744CBFD97CD62A33DEE902222F5F@VFOHMLAO13.Enterprise.afmc.ds.af.mil> Michael, Following up on a private email from a while back: > >> I remember also why I deviated from the dhcp.template > method. When I > >> put in a default entry per the man page i.e. > >> > >> cobbler system add --name=default --profile=RHEL5-x86_64 > >> > >> Followed by a > >> > >> cobbler sync > >> > >> builds dhhpd.conf with a line: hardware ethernet None; line which > >> > > dhcpd > > > >> chokes on > >> > >> > > > > I've fixed that upstream (git repo) for 0.6.1, which will probably > > release next week. If you make sure all your system entries > are named > > after a mac > > address of give an argument of --mac, that won't happen in 0.6. > > I have rebuilt the box cobbler lives on and this appears to have come back. I did repopulate the new setup with old files from /var/lib/cobbler & /etc/cobbler :-) When cobbler expands $insert_cobbler_system_definitions I get 2 host generic entries which still have the invalid entry. How best can I fix this? Thanks, Kevin From berrange at redhat.com Wed Sep 5 14:38:03 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 5 Sep 2007 15:38:03 +0100 Subject: [et-mgmt-tools] koan --virt: lvm support In-Reply-To: <1188983929.3750.7.camel@mine3.x-tend.be> References: <46DD74F5.5000506@redhat.com> <46DD928D.9090803@redhat.com> <46DD9C29.5050308@redhat.com> <20070905032841.GB8649@redhat.com> <1188983929.3750.7.camel@mine3.x-tend.be> Message-ID: <20070905143803.GA5503@redhat.com> On Wed, Sep 05, 2007 at 11:18:49AM +0200, Kris Buytaert wrote: > On Wed, 2007-09-05 at 04:28 +0100, Daniel P. Berrange wrote: > > On Tue, Sep 04, 2007 at 01:55:53PM -0400, Michael DeHaan wrote: > > > Johan Huysmans wrote: > > > >well... it is possible ;) > > > >because we are allready doing this for over a year. > > > > > > > > > > The word on the street (from Dan Berrange) is that this feature is not > > > guaranteed to work in future > > > versions of Xen. > > > > To be a little more verbose.... > > > > Virtual machines are assigned virtual *disks*. ie hdN, sdN (for full virt) or > > xvdN (for paravirt). Names like sda1 are partitions, and not disks. It makes > > no conceptual sense to have a partition without an enclosing disk. If you want > > to have a 1-1 mapping between the backing files & disks inside the guest, then > > simply don't partition the disk inside the guest - format /dev/sda directly. > > Xen paravirt only happens to allow you to map a file straight through as a > > partition, and while it may currently work, I can make absolutely no guarentees > > it will work in the future as the Xen paravirt kernel evolves - particularly > > as Xen increasingly merges into LKML trees. NB with the LKML merge the ability > > to hijack hdN and sdN in paravirt guests has been removed - paravirt can only > > use xvdN devices now. I've not checked if hijacking partitions as described > > below works, but I would not be surprised if it doesn't. Finally its also not > > portable to Xen fullyvirt, or KVM, or QEMU, etc. Basically unsupportable in > > the medium-to-long term. > > > I think it is extremely important to continue to be able to use > filesystems on LVM volumes in dom0 as filesystems in domU's. Using > them as full disks with a partition on them would be fine. Yes, that is entirely possible. If you have /dev/VolGroup00/MyGuestFS in the dom0, you can map that into the guest as /dev/xvda and simply mount /dev/xvda itself - there is no need to pretend it is /dev/xvda1 inside the guest. 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 Wed Sep 5 14:37:22 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 05 Sep 2007 10:37:22 -0400 Subject: [et-mgmt-tools] FW: Help with cobbler for dummies... In-Reply-To: <29A7C5F65D54744CBFD97CD62A33DEE902222F5F@VFOHMLAO13.Enterprise.afmc.ds.af.mil> References: <29A7C5F65D54744CBFD97CD62A33DEE902222F5F@VFOHMLAO13.Enterprise.afmc.ds.af.mil> Message-ID: <46DEBF22.1080505@redhat.com> Maute, Kevin P CTR AFRL/VAOO wrote: > > When cobbler expands $insert_cobbler_system_definitions I get 2 host > generic entries which still have the invalid entry. How best can I fix > this? > > The problem with "None" showing up in dhcpd.conf should be fixed in 0.6.1. If you are running 0.6.1 and still seeing a problem, I'd like to see the generated dhcp.conf file and what you have in /var/lib/cobbler/systems. Otherwise I'm not sure what "the invalid entry" is enough to take a guess. Let me know... Thanks! --Michael From Kevin.Maute at WPAFB.AF.MIL Wed Sep 5 15:25:37 2007 From: Kevin.Maute at WPAFB.AF.MIL (Maute, Kevin P CTR AFRL/VAOO) Date: Wed, 5 Sep 2007 11:25:37 -0400 Subject: [et-mgmt-tools] FW: Help with cobbler for dummies... In-Reply-To: <46DEBF22.1080505@redhat.com> References: <29A7C5F65D54744CBFD97CD62A33DEE902222F5F@VFOHMLAO13.Enterprise.afmc.ds.af.mil> <46DEBF22.1080505@redhat.com> Message-ID: <29A7C5F65D54744CBFD97CD62A33DEE902223014@VFOHMLAO13.Enterprise.afmc.ds.af.mil> Enjoy! > -----Original Message----- > From: et-mgmt-tools-bounces at redhat.com > [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of Michael DeHaan > Sent: Wednesday, September 05, 2007 10:37 AM > To: Fedora/Linux Management Tools > Subject: Re: [et-mgmt-tools] FW: Help with cobbler for dummies... > > Maute, Kevin P CTR AFRL/VAOO wrote: > > > > When cobbler expands $insert_cobbler_system_definitions I get 2 host > > generic entries which still have the invalid entry. How > best can I fix > > this? > > > > > The problem with "None" showing up in dhcpd.conf should be fixed in > 0.6.1. If you are running 0.6.1 and still seeing a problem, > I'd like to see > the generated dhcp.conf file and what you have in > /var/lib/cobbler/systems. Otherwise I'm not sure what "the invalid > entry" is enough to take > a guess. Let me know... > > Thanks! > > --Michael > > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > -------------- next part -------------- A non-text attachment was scrubbed... Name: dhcpd.conf Type: application/octet-stream Size: 6038 bytes Desc: dhcpd.conf URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: systems Type: application/octet-stream Size: 2705 bytes Desc: systems URL: From johan.huysmans at inuits.be Wed Sep 5 16:50:51 2007 From: johan.huysmans at inuits.be (Johan Huysmans) Date: Wed, 5 Sep 2007 18:50:51 +0200 Subject: [et-mgmt-tools] koan --virt: lvm support In-Reply-To: <20070905143803.GA5503@redhat.com> References: <46DD74F5.5000506@redhat.com> <46DD928D.9090803@redhat.com> <46DD9C29.5050308@redhat.com> <20070905032841.GB8649@redhat.com> <1188983929.3750.7.camel@mine3.x-tend.be> <20070905143803.GA5503@redhat.com> Message-ID: Manually creating these partitions is no problem, this is what that xen config does but instead of mapping it to sda1, mapping in to xvda and putting xvda in your fstab instead of sda1. But how can you tell that to koan and put that in a kickstart file. I've read in another mail on this list that it is possible to create multiple devices, but how do you tell that those devices don't need partitioning but just formatting? Johan. On 9/5/07, Daniel P. Berrange wrote: > On Wed, Sep 05, 2007 at 11:18:49AM +0200, Kris Buytaert wrote: > > On Wed, 2007-09-05 at 04:28 +0100, Daniel P. Berrange wrote: > > > On Tue, Sep 04, 2007 at 01:55:53PM -0400, Michael DeHaan wrote: > > > > Johan Huysmans wrote: > > > > >well... it is possible ;) > > > > >because we are allready doing this for over a year. > > > > > > > > > > > > > The word on the street (from Dan Berrange) is that this feature is not > > > > guaranteed to work in future > > > > versions of Xen. > > > > > > To be a little more verbose.... > > > > > > Virtual machines are assigned virtual *disks*. ie hdN, sdN (for full virt) or > > > xvdN (for paravirt). Names like sda1 are partitions, and not disks. It makes > > > no conceptual sense to have a partition without an enclosing disk. If you want > > > to have a 1-1 mapping between the backing files & disks inside the guest, then > > > simply don't partition the disk inside the guest - format /dev/sda directly. > > > Xen paravirt only happens to allow you to map a file straight through as a > > > partition, and while it may currently work, I can make absolutely no guarentees > > > it will work in the future as the Xen paravirt kernel evolves - particularly > > > as Xen increasingly merges into LKML trees. NB with the LKML merge the ability > > > to hijack hdN and sdN in paravirt guests has been removed - paravirt can only > > > use xvdN devices now. I've not checked if hijacking partitions as described > > > below works, but I would not be surprised if it doesn't. Finally its also not > > > portable to Xen fullyvirt, or KVM, or QEMU, etc. Basically unsupportable in > > > the medium-to-long term. > > > > > > I think it is extremely important to continue to be able to use > > filesystems on LVM volumes in dom0 as filesystems in domU's. Using > > them as full disks with a partition on them would be fine. > > Yes, that is entirely possible. If you have /dev/VolGroup00/MyGuestFS > in the dom0, you can map that into the guest as /dev/xvda and simply > mount /dev/xvda itself - there is no need to pretend it is /dev/xvda1 > inside the guest. > > Dan. > -- > |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| > |=- Perl modules: http://search.cpan.org/~danberr/ -=| > |=- Projects: http://freshmeat.net/~danielpb/ -=| > |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From mdehaan at redhat.com Wed Sep 5 18:10:46 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 05 Sep 2007 14:10:46 -0400 Subject: [et-mgmt-tools] FW: Help with cobbler for dummies... In-Reply-To: <29A7C5F65D54744CBFD97CD62A33DEE902223014@VFOHMLAO13.Enterprise.afmc.ds.af.mil> References: <29A7C5F65D54744CBFD97CD62A33DEE902222F5F@VFOHMLAO13.Enterprise.afmc.ds.af.mil> <46DEBF22.1080505@redhat.com> <29A7C5F65D54744CBFD97CD62A33DEE902223014@VFOHMLAO13.Enterprise.afmc.ds.af.mil> Message-ID: <46DEF126.4080802@redhat.com> Maute, Kevin P CTR AFRL/VAOO wrote: > >>> When cobbler expands $insert_cobbler_system_definitions I get 2 host >>> generic entries which still have the invalid entry. How Ugh .... still had a bug in there that can happen when the following conditions are true: (A) a system doesn't have a --mac-address set (B) the --name is not a MAC (C) manage_dhcp is turned on The fix is just: diff --git a/cobbler/action_sync.py b/cobbler/action_sync.py index 8302504..03eeb02 100644 --- a/cobbler/action_sync.py +++ b/cobbler/action_sync.py @@ -153,7 +153,7 @@ class BootSync: if mac is None or mac == "": # can't write a DHCP entry for this system # FIXME: should this be a warning? - pass + continue counter = counter + 1 systxt = "" I've checked this in upstream (git), though if you just want to apply the patch above that works also. This will also be included in 0.6.2, which I do not have a ETA on when we want to release this yet. That will probably be in a few weeks to give some time for additional testing and a few minor features to get worked in. --Michael From lfarkas at bppiac.hu Thu Sep 6 14:31:04 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Thu, 06 Sep 2007 16:31:04 +0200 Subject: [et-mgmt-tools] virt-manager install from PXE Message-ID: <46E00F28.2030207@bppiac.hu> hi, in our environment we 'd like to use kvm to test our pxe installs since all of our system is installed via pxe, so the kickstart files and all other part of the installation should have to be checked. it's currently not possible to use virt-manager to install from a PXE netboot. in qemu since 0.9 it has a boot option 'n' to boot from network. it'd useful to add such an option in virt-manager install settings (just one checkbox:-). is it require a big changes? thanks in advance. -- Levente "Si vis pacem para bellum!" From berrange at redhat.com Thu Sep 6 14:32:32 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 6 Sep 2007 15:32:32 +0100 Subject: [et-mgmt-tools] virt-manager install from PXE In-Reply-To: <46E00F28.2030207@bppiac.hu> References: <46E00F28.2030207@bppiac.hu> Message-ID: <20070906143232.GF1291@redhat.com> On Thu, Sep 06, 2007 at 04:31:04PM +0200, Farkas Levente wrote: > hi, > in our environment we 'd like to use kvm to test our pxe installs since > all of our system is installed via pxe, so the kickstart files and all > other part of the installation should have to be checked. > it's currently not possible to use virt-manager to install from a PXE > netboot. in qemu since 0.9 it has a boot option 'n' to boot from > network. it'd useful to add such an option in virt-manager install > settings (just one checkbox:-). is it require a big changes? > thanks in advance. It is a fairly simple change & one we're planning to make in the near future libvirt already supports configuring a guest to boot from network so we just need UI to enable that choice, and a fix to virt-install to allow that as an install source 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 lfarkas at bppiac.hu Thu Sep 6 14:37:50 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Thu, 06 Sep 2007 16:37:50 +0200 Subject: [et-mgmt-tools] virt-manager install from PXE In-Reply-To: <20070906143232.GF1291@redhat.com> References: <46E00F28.2030207@bppiac.hu> <20070906143232.GF1291@redhat.com> Message-ID: <46E010BE.3050706@bppiac.hu> Daniel P. Berrange wrote: > On Thu, Sep 06, 2007 at 04:31:04PM +0200, Farkas Levente wrote: >> hi, >> in our environment we 'd like to use kvm to test our pxe installs since >> all of our system is installed via pxe, so the kickstart files and all >> other part of the installation should have to be checked. >> it's currently not possible to use virt-manager to install from a PXE >> netboot. in qemu since 0.9 it has a boot option 'n' to boot from >> network. it'd useful to add such an option in virt-manager install >> settings (just one checkbox:-). is it require a big changes? >> thanks in advance. > > It is a fairly simple change & one we're planning to make in the near future > libvirt already supports configuring a guest to boot from network so we just > need UI to enable that choice, and a fix to virt-install to allow that as an > install source about when?:-) -- Levente "Si vis pacem para bellum!" From ehabkost at redhat.com Thu Sep 6 14:45:48 2007 From: ehabkost at redhat.com (Eduardo Habkost) Date: Thu, 6 Sep 2007 11:45:48 -0300 Subject: [et-mgmt-tools] [PATCH] virt-install: correct label for Fedora 8 OS variant Message-ID: <20070906144548.GI11310@blackpad.ctb.virtua.com.br> virt-manager was showing two "Fedora 7" entries. The second entry was the 'fedora8' OS variant, that had the wrong label. # HG changeset patch # User "Eduardo Habkost " # Date 1189089052 10800 # Node ID f6d797ae5cac18305fcddd57b22754020cd6f736 # Parent a6d47ed6c97cf79360ad9259554c05dcbb3f6b82 Correct label for fedora8 on os-variant list It was being shown as "Fedora 7" instead of "Fedora 8". diff -r a6d47ed6c97c -r f6d797ae5cac virtinst/FullVirtGuest.py --- a/virtinst/FullVirtGuest.py Wed Aug 29 18:25:09 2007 -0400 +++ b/virtinst/FullVirtGuest.py Thu Sep 06 11:30:52 2007 -0300 @@ -34,7 +34,7 @@ class FullVirtGuest(Guest.XenGuest): "fedora5": { "label": "Fedora Core 5", "distro": "fedora" }, \ "fedora6": { "label": "Fedora Core 6", "distro": "fedora" }, \ "fedora7": { "label": "Fedora 7", "distro": "fedora" }, \ - "fedora8": { "label": "Fedora 7", "distro": "fedora" }, \ + "fedora8": { "label": "Fedora 8", "distro": "fedora" }, \ "sles10": { "label": "Suse Linux Enterprise Server", "distro": "suse" }, \ "generic24": { "label": "Generic 2.4.x kernel" }, \ "generic26": { "label": "Generic 2.6.x kernel" }, \ -- Eduardo From mdehaan at redhat.com Thu Sep 6 14:45:53 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 06 Sep 2007 10:45:53 -0400 Subject: [et-mgmt-tools] virt-manager install from PXE In-Reply-To: <20070906143232.GF1291@redhat.com> References: <46E00F28.2030207@bppiac.hu> <20070906143232.GF1291@redhat.com> Message-ID: <46E012A1.2060102@redhat.com> Daniel P. Berrange wrote: > On Thu, Sep 06, 2007 at 04:31:04PM +0200, Farkas Levente wrote: > >> hi, >> in our environment we 'd like to use kvm to test our pxe installs since >> all of our system is installed via pxe, so the kickstart files and all >> other part of the installation should have to be checked. >> it's currently not possible to use virt-manager to install from a PXE >> netboot. in qemu since 0.9 it has a boot option 'n' to boot from >> network. it'd useful to add such an option in virt-manager install >> settings (just one checkbox:-). is it require a big changes? >> thanks in advance. >> > > It is a fairly simple change & one we're planning to make in the near future > libvirt already supports configuring a guest to boot from network so we just > need UI to enable that choice, and a fix to virt-install to allow that as an > install source > > Dan. > Just throwing a few more ideas out there -- if you're wanting to deploy Linux, in either Xen (paravirt), or qemu/kvm ... this is what koan was written for. (We don't do Xen fullvirt yet since it wants an ISO to boot.) See http://cobbler.et.redhat.com However I do agree that real PXE for virt guests would be quite useful... (and you could still use Cobbler to manage that PXE config in that case). From mdehaan at redhat.com Thu Sep 6 15:56:23 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 06 Sep 2007 11:56:23 -0400 Subject: [et-mgmt-tools] Alternative storage backends for Cobbler / other features In-Reply-To: <46D88E85.3050904@redhat.com> References: <46D88E85.3050904@redhat.com> Message-ID: <46E02327.9030500@redhat.com> Michael DeHaan wrote: > So there's been a lot of interest around having Cobbler read LDAP > recently, and possibly I'm guessing it would be useful to have it > /write/ to LDAP. I know several > folks already have their own scripts to interface between the two. > > I'm thinking about abstracting the serializer code to allow for > configs in arbitrary formats.... It's done (and available for preview now in git): https://hosted.fedoraproject.org/projects/cobbler/wiki/AlternativeSerializers There's a example idea on the Wiki page above about how you could still use Cobbler to store to disk, but get /some/ information from LDAP ... I'll leave it up to users to find creative ways to use this. I ended up /not/ writing the SQL portions (yet) as that would be a lot to maintain and it seems things like LDAP would be more useful. If extremely large numbers of profiles/nodes run us into performance issues (so far they haven't) there are probably better ways to speed up things than that initially. --Michael From rigg0022 at umn.edu Thu Sep 6 16:06:49 2007 From: rigg0022 at umn.edu (Riggs, Ben) Date: Thu, 06 Sep 2007 11:06:49 -0500 Subject: [et-mgmt-tools] PATCH: 'cobbler reposync foo bar' now syncs repos 'foo' & 'bar' regardless of 'keep-updated' flag Message-ID: <46E02599.9040509@umn.edu> Very useful for maintaining testing and production repos: Point the testing repos out and set keep updated to yes; point the production repos at /var/www/cobbler/repo_mirror/$test_repos and set keep updated to no. When the testing repos are confirmed good, 'cobbler reposync test repo names' to move test repos to production, then 'cobbler reposync' to update the test repos. From rigg0022 at umn.edu Thu Sep 6 15:54:45 2007 From: rigg0022 at umn.edu (Ben Riggs) Date: Thu, 6 Sep 2007 10:54:45 -0500 Subject: [et-mgmt-tools] [PATCH] 'cobbler reposync foo' now updates repo 'foo' whether marked to keep updated or not. Message-ID: --- cobbler/action_reposync.py | 35 +++++++++++++++++------------------ cobbler/api.py | 4 ++-- cobbler/cobbler.py | 2 +- 3 files changed, 20 insertions(+), 21 deletions(-) diff --git a/cobbler/action_reposync.py b/cobbler/action_reposync.py index 8c7e242..c8446c5 100644 --- a/cobbler/action_reposync.py +++ b/cobbler/action_reposync.py @@ -38,25 +38,34 @@ class RepoSync: """ Constructor """ - self.verbose = True - self.config = config - self.distros = config.distros() - self.profiles = config.profiles() - self.systems = config.systems() - self.settings = config.settings() - self.repos = config.repos() + self.verbose = True + self.config = config + self.distros = config.distros() + self.profiles = config.profiles() + self.systems = config.systems() + self.settings = config.settings() + self.repos = config.repos() + # ================================================================================== - def run(self,verbose=True): + def run(self, args=[], verbose=True): """ Syncs the current repo configuration file with the filesystem. """ self.verbose = verbose for repo in self.repos: + + if args != [] and repo.name not in args: + continue + elif args == [] and not repo.keep_updated: + print _("- %s is set to not be updated") % repo.name + continue + repo_path = os.path.join(self.settings.webdir, "repo_mirror", repo.name) mirror = repo.mirror + if not os.path.isdir(repo_path) and not repo.mirror.lower().startswith("rhn://"): os.makedirs(repo_path) @@ -92,13 +101,6 @@ class RepoSync: if repo.rpm_list != "": has_rpm_list = True - # user might have disabled repo updates in the config file for whatever reason. - # if so, don't update this one. - - if not repo.keep_updated: - print _("- %s is set to not be updated") % repo.name - return True - # create yum config file for use by reposync store_path = os.path.join(self.settings.webdir, "repo_mirror") dest_path = os.path.join(store_path, repo.name) @@ -188,9 +190,6 @@ class RepoSync: Handle copying of rsync:// and rsync-over-ssh repos. """ - if not repo.keep_updated: - print _("- %s is set to not be updated") % repo.name - return True if repo.rpm_list != "": print _("- warning: --rpm-list is not supported for rsync'd repositories") dest_path = os.path.join(self.settings.webdir, "repo_mirror", repo.name) diff --git a/cobbler/api.py b/cobbler/api.py index f5d3322..304eddb 100644 --- a/cobbler/api.py +++ b/cobbler/api.py @@ -188,13 +188,13 @@ class BootAPI: sync = action_sync.BootSync(self._config) return sync.run() - def reposync(self): + def reposync(self, args=[]): """ Take the contents of /var/lib/cobbler/repos and update them -- or create the initial copy if no contents exist yet. """ reposync = action_reposync.RepoSync(self._config) - return reposync.run() + return reposync.run(args) def status(self,mode): statusifier = action_status.BootStatusReport(self._config, mode) diff --git a/cobbler/cobbler.py b/cobbler/cobbler.py index 7ecb2d6..9d166e3 100755 --- a/cobbler/cobbler.py +++ b/cobbler/cobbler.py @@ -581,7 +581,7 @@ class BootCLI: Sync the repo-specific portions of the config with the filesystem. 'cobbler reposync'. Intended to be run on cron. """ - self.api.reposync() + self.api.reposync(args) return True def validateks(self,args): -- 1.5.2.4 --------------000401020305000101040708-- From chris.sarginson at ukfast.net Thu Sep 6 16:46:31 2007 From: chris.sarginson at ukfast.net (Chris Sarginson) Date: Thu, 06 Sep 2007 17:46:31 +0100 Subject: [et-mgmt-tools] Virt-install query Message-ID: <46E02EE7.10509@ukfast.net> Hi Chaps, I'm currently running CentOS 5 with python-virtinst-0.99.0-2.el5.centos, and all I am looking to do is set acpi/apic/pae to be 0, so that I can install windows - with these set to 1 the setup wont continue. Is there are an argument I can provide to virt-install to change these? I have tried the following, but as the /etc/xen/logan file contains acpi=1 apic=1 pae=1 It's not working. Command I am using currently: virt-install -n logan -r 512 -vcpus=1 -f /home/vm/102 -s 10 -v --cdrom=/root/standard2003.iso --noautoconsole --vnc -- 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 mdehaan at redhat.com Thu Sep 6 16:52:55 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 06 Sep 2007 12:52:55 -0400 Subject: [et-mgmt-tools] PATCH: 'cobbler reposync foo bar' now syncs repos 'foo' & 'bar' regardless of 'keep-updated' flag In-Reply-To: <46E02599.9040509@umn.edu> References: <46E02599.9040509@umn.edu> Message-ID: <46E03067.4040604@redhat.com> Riggs, Ben wrote: > Very useful for maintaining testing and production repos: Point the > testing repos out and set keep updated to yes; point the production > repos at /var/www/cobbler/repo_mirror/$test_repos and set keep updated > to no. When the testing repos are confirmed good, 'cobbler reposync > test repo names' to move test repos to production, then 'cobbler > reposync' to update the test repos. > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools Applied and very useful, thanks! By "cobbler reposync test repo names", Ben means that you can list repos that you want to explictly sync on the command line, without having to sync everything. There isn't a command called "test repo names" :) As an example, basically "cobbler reposync fc6i386updates fc6x8664updates" would sync both sets of updates for fc6, and leave out f7. You can sync as few or as many repos as you like. If you only want to edit a certain repo sporadically (say a repo for 3rd party drivers or ISV software, or a mirror of some RPMs your company has produced internally), "repo edit --name=foo --keep-updated=False" will do the trick, and "cobbler reposync foo" will still force it to sync when needed. --Michael From Kevin.Maute at WPAFB.AF.MIL Thu Sep 6 17:49:38 2007 From: Kevin.Maute at WPAFB.AF.MIL (Maute, Kevin P CTR AFRL/VAOO) Date: Thu, 6 Sep 2007 13:49:38 -0400 Subject: [et-mgmt-tools] Next challenge - anaconda confusion Message-ID: <29A7C5F65D54744CBFD97CD62A33DEE9022239CF@VFOHMLAO13.Enterprise.afmc.ds.af.mil> Folks, I am still struggling to get back to what I had earlier... Now I am getting the following error: Unable to read package metadata. This may br due to a missing repodata directory. Please ensure that your install tree has been correctly generated. Cannot open/read repod.xml file for repository: anaconda-base-200702081643.x86_64 I have even rebuilt the RHEL5 Client repo with no change so am I missing a createrepo somewhere or is this a known bug with cobbler (0.6.1)? I am also including a 'cobbler report' in case that helps... The repos were a last ditch attempt to get this working, symptoms did not change with/without them. Thanks, Kevin <> -------------------------------------------- Kevin Maute Sr. Linux/Unix System Administrator RCF Information Systems AFRL/VAOO Contractor 937.255.1810 kevin.maute at wpafb.af.mil -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: cobbler-report-9-6 Type: application/octet-stream Size: 1329 bytes Desc: cobbler-report-9-6 URL: From mdehaan at redhat.com Thu Sep 6 19:31:05 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 06 Sep 2007 15:31:05 -0400 Subject: [et-mgmt-tools] Next challenge - anaconda confusion In-Reply-To: <29A7C5F65D54744CBFD97CD62A33DEE9022239CF@VFOHMLAO13.Enterprise.afmc.ds.af.mil> References: <29A7C5F65D54744CBFD97CD62A33DEE9022239CF@VFOHMLAO13.Enterprise.afmc.ds.af.mil> Message-ID: <46E05579.7000200@redhat.com> Maute, Kevin P CTR AFRL/VAOO wrote: > > Folks, > > I am still struggling to get back to what I had earlier? Now I am > getting the following error: > > Unable to read package metadata. This may br due to a missing repodata > directory. Please ensure that your install tree has been correctly > generated. Cannot open/read repod.xml file for repository: > anaconda-base-200702081643.x86_64 > > I have even rebuilt the RHEL5 Client repo with no change so am I > missing a createrepo somewhere or is this a known bug with cobbler > (0.6.1)? I am also including a 'cobbler report' in case that helps? > > There isn't a Cobbler bug as we have several folks using the repo management features with 0.6.1 ... there's not quite enough information to say what the actual case is. I hcan't quite tell /which/ repo you are talking about, what the source for that repo is, or the parameters that created it. If you are still talking about the manual "cobbler repo adds" you used to create the RHEL5 install, you might want to reimport the distro (cobbler import -- and yes, you can import over an existing one) and it will set them up automatically as "source" repos, which are static and don't need to be updated -- therefore you don't need to use "cobbler repo add" to set them up (you can "cobbler repo remove" the ones you have and clean out their storage in /var/www/cobbler/repo_mirror). As I recall you had added them manually, and those are the repos you used (stuff off the install CD), that's the way they are designed to work for distros that have the split layout (like RHEL5). If this is instead talking about an updates repository, etc, disregard the above. --Michael From alikins at redhat.com Thu Sep 6 22:04:13 2007 From: alikins at redhat.com (Adrian Likins) Date: Thu, 06 Sep 2007 18:04:13 -0400 Subject: [et-mgmt-tools] virt-factory 0.0.4 released Message-ID: <46E0795D.9060606@redhat.com> Virt-factory 0.0.4 has been released. Virt-Factory is a set of applications for managing a very large number of virtual machines such as might be found in data centers or compute farms. The goal of the project is to determine new and efficient ways to control these large environments. Changes include: * Use busrpc(AMQP) to communicate with nodes * Status tracking -- heartbeat from node daemons and associated infrastructure. async notification of changes to Xen state show up in WUI, use AMQP to do this. * Keeping VM's (re)started when host nodes restart. * Separate WUI from central server so they can optionally be installed on separate machines. * new command line app "ampm" that has same functionality as the wui * virt-factory now supports KVM for virt, and uses it by default. For now, F7 is required (AMQP being the main reason) Our plans for future releases are detailed here: http://virt-factory.et.redhat.com/vf-roadmap.php More info, including download info, etc can be found at: http://virt-factory.et.redhat.com/ It's still in early stages, but it's getting there. Downloads are here: http://virt-factory.et.redhat.com/download.php Thanks! Michael DeHaan Adrian Likins Scott Seago From smooge at gmail.com Thu Sep 6 22:43:48 2007 From: smooge at gmail.com (Stephen John Smoogen) Date: Thu, 6 Sep 2007 16:43:48 -0600 Subject: [et-mgmt-tools] virt-factory 0.0.4 released In-Reply-To: <46E0795D.9060606@redhat.com> References: <46E0795D.9060606@redhat.com> Message-ID: <80d7e4090709061543s149a4744i3ea5d8983fe80c5b@mail.gmail.com> On 9/6/07, Adrian Likins wrote: > > Virt-factory 0.0.4 has been released. > > Virt-Factory is a set of applications for managing a very large number > of virtual machines such as might be found in data centers or compute > farms. The goal of the project is to determine new and efficient ways to > control these large environments. > > Changes include: > > * Use busrpc(AMQP) to communicate with nodes > For now, F7 is required (AMQP being the main reason) > > So this will be more of a RHEL-6 possible product than an EL-5 one? I am just working out which items I can experiment with here -- Stephen J Smoogen. -- CSIRT/Linux System Administrator How far that little candle throws his beams! So shines a good deed in a naughty world. = Shakespeare. "The Merchant of Venice" From mdehaan at redhat.com Thu Sep 6 23:14:57 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 06 Sep 2007 19:14:57 -0400 Subject: [et-mgmt-tools] virt-factory 0.0.4 released In-Reply-To: <80d7e4090709061543s149a4744i3ea5d8983fe80c5b@mail.gmail.com> References: <46E0795D.9060606@redhat.com> <80d7e4090709061543s149a4744i3ea5d8983fe80c5b@mail.gmail.com> Message-ID: <46E089F1.1050708@redhat.com> Stephen John Smoogen wrote: > On 9/6/07, Adrian Likins wrote: > >> Virt-factory 0.0.4 has been released. >> >> Virt-Factory is a set of applications for managing a very large number >> of virtual machines such as might be found in data centers or compute >> farms. The goal of the project is to determine new and efficient ways to >> control these large environments. >> >> Changes include: >> >> * Use busrpc(AMQP) to communicate with nodes >> > > >> For now, F7 is required (AMQP being the main reason) >> >> >> > > So this will be more of a RHEL-6 possible product than an EL-5 one? I > am just working out which items I can experiment with here > > For the next release we are going to be looking harder at supporting other distros as an install base -- RHEL-5 is pretty much that other OS -- this means EL5 RPMs, AMQP, and much easier ways to add new distros to the Virt-Factory DB over time. For the most part, we expect AMQP to be available for those distros soon enough -- It's already in Fedora, so that's why we are using F-7 now. There aren't any other pieces that are really keeping us from doing RHEL5 soon, at least in the area of Xen support. --Michael From karl at Clickability.com Fri Sep 7 00:21:29 2007 From: karl at Clickability.com (Karl Balsmeier) Date: Thu, 6 Sep 2007 17:21:29 -0700 Subject: [et-mgmt-tools] cobbler import - centos 4.5/64 rsync mirror path Message-ID: <4DC963D8A360F243BE64E2A3F68205A2018CCA03@webmail.clickability.com> Hi, In an attempt to grab a distro via rsync, I am doing the following command, substituting the servergoeshere for the real rsync mirror. The problem is that I have no way to navigate down the tree far enough to actually make this work, because obviously my browser doesn't allow this protocol. cobbler import --mirror=rsync://servergoeshere/path/to/distro --name=Centos45-64 I visit: http://www.centos.org/modules/tinycontent/index.php?id=13 and pick out a mirror: rsync://mirrors.kernel.org/centos/ but can't seem to give cobbler the proper path to the distro. I navigated the http dir structure and tried to guess, but no dice. What do you do when looking for the right path to give cobbler to a distro in rsync? Do you use the rsync client? is there a smart and useful command out there that reveals the right path to a distro on an rsync mirror? -krb -------------- next part -------------- An HTML attachment was scrubbed... URL: From wright at imageworks.com Fri Sep 7 00:27:26 2007 From: wright at imageworks.com (Peter Wright) Date: Thu, 06 Sep 2007 17:27:26 -0700 Subject: [et-mgmt-tools] cobbler import - centos 4.5/64 rsync mirror path In-Reply-To: <4DC963D8A360F243BE64E2A3F68205A2018CCA03@webmail.clickability.com> References: <4DC963D8A360F243BE64E2A3F68205A2018CCA03@webmail.clickability.com> Message-ID: <46E09AEE.3060000@imageworks.com> Karl Balsmeier wrote: > > Hi, > > In an attempt to grab a distro via rsync, I am doing the following > command, substituting the servergoeshere for the real rsync mirror. > > The problem is that I have no way to navigate down the tree far enough > to actually make this work, because obviously my browser doesn't allow > this protocol. > > cobbler import --mirror=rsync://servergoeshere/path/to/distro > --name=Centos45-64 > > I visit: http://www.centos.org/modules/tinycontent/index.php?id=13 > > and pick out a mirror: rsync://mirrors.kernel.org/centos/ > > but can't seem to give cobbler the proper path to the distro. I > navigated the http dir structure and tried to guess, but no dice. > > What do you do when looking for the right path to give cobbler to a > distro in rsync? > > Do you use the rsync client? is there a smart and useful command out > there that reveals the right path to a distro on an rsync mirror? > > -krb > you can usually just FTP to these mirrors and grab the appropriate path from there. in this case you can actually goto mirrors.kernel.org/centos via http as well. then drill down to the branch that you want to sync and you should be able to use that path for your mirror switch. -pete -- Peter Wright Systems Administrator Sony Pictures Imageworks wright at imageworks.com www.imageworks.com From karl at Clickability.com Fri Sep 7 00:29:48 2007 From: karl at Clickability.com (Karl Balsmeier) Date: Thu, 6 Sep 2007 17:29:48 -0700 Subject: [et-mgmt-tools] cobbler import - centos 4.5/64 rsync mirror path References: <4DC963D8A360F243BE64E2A3F68205A2018CCA03@webmail.clickability.com> <46E09AEE.3060000@imageworks.com> Message-ID: <4DC963D8A360F243BE64E2A3F68205A2018CCA04@webmail.clickability.com> yep, one would think this would work out great, but there's perhaps a clarification needed, -the distro dir should look like this in the end result right?: RPMS base or should I be higher up? -krb -----Original Message----- From: et-mgmt-tools-bounces at redhat.com on behalf of Peter Wright Sent: Thu 9/6/2007 5:27 PM To: Fedora/Linux Management Tools Subject: Re: [et-mgmt-tools] cobbler import - centos 4.5/64 rsync mirror path Karl Balsmeier wrote: > > Hi, > > In an attempt to grab a distro via rsync, I am doing the following > command, substituting the servergoeshere for the real rsync mirror. > > The problem is that I have no way to navigate down the tree far enough > to actually make this work, because obviously my browser doesn't allow > this protocol. > > cobbler import --mirror=rsync://servergoeshere/path/to/distro > --name=Centos45-64 > > I visit: http://www.centos.org/modules/tinycontent/index.php?id=13 > > and pick out a mirror: rsync://mirrors.kernel.org/centos/ > > but can't seem to give cobbler the proper path to the distro. I > navigated the http dir structure and tried to guess, but no dice. > > What do you do when looking for the right path to give cobbler to a > distro in rsync? > > Do you use the rsync client? is there a smart and useful command out > there that reveals the right path to a distro on an rsync mirror? > > -krb > you can usually just FTP to these mirrors and grab the appropriate path from there. in this case you can actually goto mirrors.kernel.org/centos via http as well. then drill down to the branch that you want to sync and you should be able to use that path for your mirror switch. -pete -- Peter Wright Systems Administrator Sony Pictures Imageworks wright at imageworks.com www.imageworks.com _______________________________________________ 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 karl at Clickability.com Fri Sep 7 00:35:20 2007 From: karl at Clickability.com (Karl Balsmeier) Date: Thu, 6 Sep 2007 17:35:20 -0700 Subject: success! RE: [et-mgmt-tools] cobbler import - centos 4.5/64 rsync References: <4DC963D8A360F243BE64E2A3F68205A2018CCA03@webmail.clickability.com> <46E09AEE.3060000@imageworks.com> Message-ID: <4DC963D8A360F243BE64E2A3F68205A2018CCA05@webmail.clickability.com> I got it to work with: cobbler import --mirror=rsync://mirrors.kernel.org/centos/4.5/os/x86_64/CentOS/ --name=C45-64 I did get this error below, but I think that's because I removed the ia64 statement from /var/lib/cobbler/settings: (sound correct?) Traceback (most recent call last): File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 636, in main BootCLI(sys.argv).run() File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 110, in run self.curry_args(self.args[1:], self.commands['toplevel']) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 465, in curry_args commands[args[0]](args[1:]) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 588, in import_tree return self.apply_args(args,commands,on_ok) File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 453, in apply_args on_ok() File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 587, in on_ok = lambda: go_import() File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 580, in go_import self.temp_mirror_name) File "/usr/lib/python2.4/site-packages/cobbler/api.py", line 159, in import_tree return importer.run() File "/usr/lib/python2.4/site-packages/cobbler/action_import.py", line 92, in run self.api.sync() File "/usr/lib/python2.4/site-packages/cobbler/api.py", line 129, in sync return sync.run() File "/usr/lib/python2.4/site-packages/cobbler/action_sync.py", line 70, in run self.write_dhcp_file() File "/usr/lib/python2.4/site-packages/cobbler/action_sync.py", line 132, in write_dhcp_file elilo = os.path.basename(self.settings.bootloaders["ia64"]) KeyError: 'ia64' -krb -------------- next part -------------- An HTML attachment was scrubbed... URL: From lfarkas at bppiac.hu Fri Sep 7 09:39:35 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Fri, 07 Sep 2007 11:39:35 +0200 Subject: [et-mgmt-tools] bug in virt-manager with bridged kvm Message-ID: <46E11C57.8020607@bppiac.hu> hi, when i try to install a new kvm guest with brided eth0 i've got the following error (and can't install the guest): -------------------------------------------- libvir: QEMU error : TAP bridge path 'peth0 (Bridge eth0)' is too long -------------------------------------------- previously it was only 'eth0' and not 'peth0 (Bridge eth0)'. -- Levente "Si vis pacem para bellum!" From marcmo at foxriver.com Fri Sep 7 12:54:02 2007 From: marcmo at foxriver.com (Marc Mondragon) Date: Fri, 7 Sep 2007 07:54:02 -0500 Subject: [et-mgmt-tools] Cobbler 0.6.2-1 error Message-ID: All- I was interested in trying out the new reposync stuff and after I grabbed the source from git, rebuilt it and ran a command I got this: %cobbler sync Traceback (most recent call last): File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 711, in main BootCLI(sys.argv).run() File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 42, in __init__ self.api = api.BootAPI() File "/usr/lib/python2.4/site-packages/cobbler/api.py", line 43, in __init__ self.deserialize() File "/usr/lib/python2.4/site-packages/cobbler/api.py", line 227, in deserialize return self._config.deserialize() File "/usr/lib/python2.4/site-packages/cobbler/config.py", line 165, in deserialize if not serializer.deserialize(x,topological=True): File "/usr/lib/python2.4/site-packages/cobbler/serializer.py", line 45, in deserialize return storage_module.deserialize(obj,topological) File "/usr/lib/python2.4/site-packages/cobbler/modules/serializer_yaml.py", line 92, in deserialize obj.from_datastruct(datastruct) File "/usr/lib/python2.4/site-packages/cobbler/collection.py", line 137, in from_datastruct item = self.factory_produce(self.config,seed_data) File "/usr/lib/python2.4/site-packages/cobbler/collection_repos.py", line 37, in factory_produce return repo.Repo(config).from_datastruct(seed_data) File "/usr/lib/python2.4/site-packages/cobbler/item_repo.py", line 52, in from_datastruct self.set_keep_updated(self.keep_updated) File "/usr/lib/python2.4/site-packages/cobbler/item_repo.py", line 79, in set_keep_updated if not keep_updated.lower() in ["yes","y","yup","yeah","1","true"]: AttributeError: 'int' object has no attribute 'lower' Marc Mondragon Fox River Financial Resources/Ritchie Capital Investments, Ltd. 2100 Enterprise Avenue Geneva, IL 60134 marcmo at foxriver.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From marcmo at foxriver.com Fri Sep 7 13:11:46 2007 From: marcmo at foxriver.com (Marc Mondragon) Date: Fri, 7 Sep 2007 08:11:46 -0500 Subject: [et-mgmt-tools] RE: Cobbler 0.6.2-1 error In-Reply-To: References: Message-ID: Sorry to be replying to my own message ... If I change all the keep_updated to 'y' from '1' in /var/lib/cobbler/repos the error I was seeing goes away. Marc Mondragon Fox River Financial Resources/Ritchie Capital Investments, Ltd. 2100 Enterprise Avenue Geneva, IL 60134 marcmo at foxriver.com ________________________________ From: Marc Mondragon Sent: Friday, September 07, 2007 7:54 AM To: 'et-mgmt-tools at redhat.com' Subject: Cobbler 0.6.2-1 error All- I was interested in trying out the new reposync stuff and after I grabbed the source from git, rebuilt it and ran a command I got this: %cobbler sync Traceback (most recent call last): File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 711, in main BootCLI(sys.argv).run() File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 42, in __init__ self.api = api.BootAPI() File "/usr/lib/python2.4/site-packages/cobbler/api.py", line 43, in __init__ self.deserialize() File "/usr/lib/python2.4/site-packages/cobbler/api.py", line 227, in deserialize return self._config.deserialize() File "/usr/lib/python2.4/site-packages/cobbler/config.py", line 165, in deserialize if not serializer.deserialize(x,topological=True): File "/usr/lib/python2.4/site-packages/cobbler/serializer.py", line 45, in deserialize return storage_module.deserialize(obj,topological) File "/usr/lib/python2.4/site-packages/cobbler/modules/serializer_yaml.py", line 92, in deserialize obj.from_datastruct(datastruct) File "/usr/lib/python2.4/site-packages/cobbler/collection.py", line 137, in from_datastruct item = self.factory_produce(self.config,seed_data) File "/usr/lib/python2.4/site-packages/cobbler/collection_repos.py", line 37, in factory_produce return repo.Repo(config).from_datastruct(seed_data) File "/usr/lib/python2.4/site-packages/cobbler/item_repo.py", line 52, in from_datastruct self.set_keep_updated(self.keep_updated) File "/usr/lib/python2.4/site-packages/cobbler/item_repo.py", line 79, in set_keep_updated if not keep_updated.lower() in ["yes","y","yup","yeah","1","true"]: AttributeError: 'int' object has no attribute 'lower' Marc Mondragon Fox River Financial Resources/Ritchie Capital Investments, Ltd. 2100 Enterprise Avenue Geneva, IL 60134 marcmo at foxriver.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From berrange at redhat.com Fri Sep 7 13:48:01 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Fri, 7 Sep 2007 14:48:01 +0100 Subject: [et-mgmt-tools] bug in virt-manager with bridged kvm In-Reply-To: <46E11C57.8020607@bppiac.hu> References: <46E11C57.8020607@bppiac.hu> Message-ID: <20070907134801.GA10078@redhat.com> On Fri, Sep 07, 2007 at 11:39:35AM +0200, Farkas Levente wrote: > hi, > when i try to install a new kvm guest with brided eth0 i've got the > following error (and can't install the guest): > -------------------------------------------- > libvir: QEMU error : TAP bridge path 'peth0 (Bridge eth0)' is too long > -------------------------------------------- > previously it was only 'eth0' and not 'peth0 (Bridge eth0)'. Yes that's a bug in 0.5.0 which I just got a patch for yesterday: http://hg.et.redhat.com/virt/applications/virt-manager--devel?cs=a5b0353ac6e1 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 Sep 7 13:53:44 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 07 Sep 2007 09:53:44 -0400 Subject: success! RE: [et-mgmt-tools] cobbler import - centos 4.5/64 rsync In-Reply-To: <4DC963D8A360F243BE64E2A3F68205A2018CCA05@webmail.clickability.com> References: <4DC963D8A360F243BE64E2A3F68205A2018CCA03@webmail.clickability.com> <46E09AEE.3060000@imageworks.com> <4DC963D8A360F243BE64E2A3F68205A2018CCA05@webmail.clickability.com> Message-ID: <46E157E8.40707@redhat.com> Karl Balsmeier wrote: > > I got it to work with: > > cobbler import > --mirror=rsync://mirrors.kernel.org/centos/4.5/os/x86_64/CentOS/ > --name=C45-64 > > I did get this error below, but I think that's because I removed the > ia64 statement from /var/lib/cobbler/settings: > (sound correct?) > You shouldn't have removed the IA64 statement. Cobbler ships elilo as part of it's config, so the file is present and won't hurt anything unless you have systems tagged as "ia64". Put it back :) bootloaders: ia64: /var/lib/cobbler/elilo-3.6-ia64.efi standard: /usr/lib/syslinux/pxelinux.0 From mdehaan at redhat.com Fri Sep 7 13:56:31 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 07 Sep 2007 09:56:31 -0400 Subject: [et-mgmt-tools] RE: Cobbler 0.6.2-1 error In-Reply-To: References: Message-ID: <46E1588F.2050206@redhat.com> Marc Mondragon wrote: > > Sorry to be replying to my own message ? > > If I change all the keep_updated to ?y? from ?1? in > /var/lib/cobbler/repos the error I was seeing goes away. > > Marc Mondragon > Thanks for the report! I'll put in a quick fix shortly :) --Michael > Fox River Financial Resources/Ritchie Capital Investments, Ltd. > 2100 Enterprise Avenue > Geneva, IL 60134 > marcmo at foxriver.com > > ------------------------------------------------------------------------ > > *From:* Marc Mondragon > *Sent:* Friday, September 07, 2007 7:54 AM > *To:* 'et-mgmt-tools at redhat.com' > *Subject:* Cobbler 0.6.2-1 error > > All- > > I was interested in trying out the new reposync stuff and after I > grabbed the source from git, > > rebuilt it and ran a command I got this: > > %cobbler sync > > Traceback (most recent call last): > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 711, > in main > > BootCLI(sys.argv).run() > > File "/usr/lib/python2.4/site-packages/cobbler/cobbler.py", line 42, > in __init__ > > self.api = api.BootAPI() > > File "/usr/lib/python2.4/site-packages/cobbler/api.py", line 43, in > __init__ > > self.deserialize() > > File "/usr/lib/python2.4/site-packages/cobbler/api.py", line 227, in > deserialize > > return self._config.deserialize() > > File "/usr/lib/python2.4/site-packages/cobbler/config.py", line 165, > in deserialize > > if not serializer.deserialize(x,topological=True): > > File "/usr/lib/python2.4/site-packages/cobbler/serializer.py", line > 45, in deserialize > > return storage_module.deserialize(obj,topological) > > File > "/usr/lib/python2.4/site-packages/cobbler/modules/serializer_yaml.py", > line 92, in deserialize > > obj.from_datastruct(datastruct) > > File "/usr/lib/python2.4/site-packages/cobbler/collection.py", line > 137, in from_datastruct > > item = self.factory_produce(self.config,seed_data) > > File "/usr/lib/python2.4/site-packages/cobbler/collection_repos.py", > line 37, in factory_produce > > return repo.Repo(config).from_datastruct(seed_data) > > File "/usr/lib/python2.4/site-packages/cobbler/item_repo.py", line 52, > in from_datastruct > > self.set_keep_updated(self.keep_updated) > > File "/usr/lib/python2.4/site-packages/cobbler/item_repo.py", line 79, > in set_keep_updated > > if not keep_updated.lower() in ["yes","y","yup","yeah","1","true"]: > > AttributeError: 'int' object has no attribute 'lower' > > Marc Mondragon > > Fox River Financial Resources/Ritchie Capital Investments, Ltd. > 2100 Enterprise Avenue > Geneva, IL 60134 > marcmo at foxriver.com > > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools From mail-lists at karan.org Fri Sep 7 16:52:39 2007 From: mail-lists at karan.org (Karanbir Singh) Date: Fri, 07 Sep 2007 17:52:39 +0100 Subject: [et-mgmt-tools] virt-factory 0.0.4 released In-Reply-To: <46E089F1.1050708@redhat.com> References: <46E0795D.9060606@redhat.com> <80d7e4090709061543s149a4744i3ea5d8983fe80c5b@mail.gmail.com> <46E089F1.1050708@redhat.com> Message-ID: <46E181D7.6090104@karan.org> Michael DeHaan wrote: > For the most part, we expect AMQP to be available for those distros soon > enough -- It's already in Fedora, so that's why we are using F-7 now. Just wondering if there was anything that we could do, to help this process along and get the required bits setup for CentOS/RHEL-5 -- Karanbir Singh : http://www.karan.org/ : 2522219 at icq From mdehaan at redhat.com Fri Sep 7 17:00:23 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 07 Sep 2007 13:00:23 -0400 Subject: [et-mgmt-tools] virt-factory 0.0.4 released In-Reply-To: <46E181D7.6090104@karan.org> References: <46E0795D.9060606@redhat.com> <80d7e4090709061543s149a4744i3ea5d8983fe80c5b@mail.gmail.com> <46E089F1.1050708@redhat.com> <46E181D7.6090104@karan.org> Message-ID: <46E183A7.7070009@redhat.com> Karanbir Singh wrote: > Michael DeHaan wrote: >> For the most part, we expect AMQP to be available for those distros >> soon enough -- It's already in Fedora, so that's why we are using F-7 >> now. > > Just wondering if there was anything that we could do, to help this > process along and get the required bits setup for CentOS/RHEL-5 > Might want to ask that question to rhm-users at redhat.com --Michael From smooge at gmail.com Fri Sep 7 17:42:31 2007 From: smooge at gmail.com (Stephen John Smoogen) Date: Fri, 7 Sep 2007 11:42:31 -0600 Subject: [et-mgmt-tools] virt-factory 0.0.4 released In-Reply-To: <46E183A7.7070009@redhat.com> References: <46E0795D.9060606@redhat.com> <80d7e4090709061543s149a4744i3ea5d8983fe80c5b@mail.gmail.com> <46E089F1.1050708@redhat.com> <46E181D7.6090104@karan.org> <46E183A7.7070009@redhat.com> Message-ID: <80d7e4090709071042y3b7ba0d3u9b94b70c6718097f@mail.gmail.com> On 9/7/07, Michael DeHaan wrote: > Karanbir Singh wrote: > > Michael DeHaan wrote: > >> For the most part, we expect AMQP to be available for those distros > >> soon enough -- It's already in Fedora, so that's why we are using F-7 > >> now. > > > > Just wondering if there was anything that we could do, to help this > > process along and get the required bits setup for CentOS/RHEL-5 > > > > Might want to ask that question to rhm-users at redhat.com > What is rhm-users? I havent seen any announce for it.. it has no description and it has 1 post which was more of a shotgun email from someone trying to get support 2-3 months ago. -- Stephen J Smoogen. -- CSIRT/Linux System Administrator How far that little candle throws his beams! So shines a good deed in a naughty world. = Shakespeare. "The Merchant of Venice" From mdehaan at redhat.com Fri Sep 7 19:22:14 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 07 Sep 2007 15:22:14 -0400 Subject: [et-mgmt-tools] virt-factory 0.0.4 released In-Reply-To: <80d7e4090709071042y3b7ba0d3u9b94b70c6718097f@mail.gmail.com> References: <46E0795D.9060606@redhat.com> <80d7e4090709061543s149a4744i3ea5d8983fe80c5b@mail.gmail.com> <46E089F1.1050708@redhat.com> <46E181D7.6090104@karan.org> <46E183A7.7070009@redhat.com> <80d7e4090709071042y3b7ba0d3u9b94b70c6718097f@mail.gmail.com> Message-ID: <46E1A4E6.503@redhat.com> Stephen John Smoogen wrote: > On 9/7/07, Michael DeHaan wrote: > >> Karanbir Singh wrote: >> >>> Michael DeHaan wrote: >>> >>>> For the most part, we expect AMQP to be available for those distros >>>> soon enough -- It's already in Fedora, so that's why we are using F-7 >>>> now. >>>> >>> Just wondering if there was anything that we could do, to help this >>> process along and get the required bits setup for CentOS/RHEL-5 >>> >>> >> Might want to ask that question to rhm-users at redhat.com >> >> > > What is rhm-users? I havent seen any announce for it.. it has no > description and it has 1 post which was more of a shotgun email from > someone trying to get support 2-3 months ago. > > Didn't know that, the QPID folks suggested I give out that address -- there is also a development list at qpid-dev at incubator.apache.org that seems to be pretty active. --Michael From mdehaan at redhat.com Fri Sep 7 20:49:07 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 07 Sep 2007 16:49:07 -0400 Subject: [et-mgmt-tools] How to use Cobbler to manage DHCP automatically when you have multiple subnets Message-ID: <46E1B943.8080203@redhat.com> Discussion on IRC today yielded the need for a new feature, which you can try out in the latest git. The feature is this. Make Cobbler's integrated DHCP management work better with configurations that require multiple subnets. Originally I wrote dhcp management as a helper for "simple" deployment cases. This allows it to be used in much more complex deployments. So I think this is pretty neat: New option for "system add" commands: --dhcp-tag If you are setting up a PXE environment with multiple subnets/gateways, and are using cobbler to manage a DHCP configuration, you will want this option. By default, the dhcp tag for all systems is "default" and means that in the DHCP template files the systems will expand out where $insert_cobbler_systems_definitions is found in the DHCP template. However, you may want certain systems to expand out in other places in the file. Setting --dhcp-tag=subnet2 for instance, will cause that system to expand out where $insert_cobbler_system_definitions_subnet2 is found, allowing you to insert directives to specify different subnets (or other parameters) before the DHCP configuration entries for those particular systems. Example: cobbler system add --name=foo --mac=AA:BB:CC:DD:EE:CC --dhcp-tag=megalab Example: cobbler system add --name=foo --mac=AA:BB:CC:DD:EE:DD --dhcp-tag=superlab Example: cobbler system add --name=foo --mac=AA:BB:CC:DD:EE:EE Example: cobbler system add --name=foo --mac=AA:BB:CC:DD:EE:FF In the above examples, when you look at /etc/cobbler/dhcp.template, the last two systems would expand out where you see "$insert_cobbler_system_definitions". The other two, respectively, would expand out with "$insert_cobbler_system_definitions_megalab" and "$insert_cobbler_system_definitions_superlab". This allows you to assign different DHCP parameters to "superlab" and "megalab" ... and prevents you from having to run multiple DHCP servers or turn off cobbler's built in DHCP management features. (As an aside: You can also specify --dhcp-tag on the "profile" object, for convience ... so you might have a profile called "fc6webservers" and another profile that inherits from that called "fc6webservers-in-megalab" ... and systems added to that profile would work automatically.) --Michael From mdehaan at redhat.com Fri Sep 7 20:56:08 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 07 Sep 2007 16:56:08 -0400 Subject: [et-mgmt-tools] Cobbler, koan, and XMLRPC across firewall boundaries made easier Message-ID: <46E1BAE8.4070307@redhat.com> Hi folks, I've put in some code into 0.6.2 to configure Apache for doing a Reverse proxy on Cobbler's XMLRPC port -- and I've modified koan to be able to know about it. What does this mean? This means you no longer have to open up tcp port 25151 on your firewall to let koan through. It can just use the commonly unblocked port 80 and your other network admins/users don't have to know any different. The port 80 XMLRPC server endpoint we are using is: http://server:80/cobbler_api (As I was discussing earlier, we'll probably do similar for our pending read-write XMLRPC API ... using Apache for access control/security at least at first. The read_write API will also be off by default -- that endpoint will likely be http(s)://server/cobbler_api_rw) If you don't modify your /etc/httpd/conf.d/cobbler.conf, koan will still know how to use the 25151 port. If you install the RPM new, you'll get the config file set up for the Reverse Proxy. If you have an existing install and want this behavior, just copy the rpmsave file over the /etc/httpd/conf.d/cobbler.conf and restart Apache. There you go... koan over port 80. --Michael From lfarkas at bppiac.hu Sat Sep 8 11:15:04 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Sat, 08 Sep 2007 13:15:04 +0200 Subject: [et-mgmt-tools] (long) kvm and virt-manager not ready for daily usage Message-ID: <46E28438.5000403@bppiac.hu> hi, in the last 2 weeks we play a lot with our new server which we but to a our virtual server for the development and collect some very subjective experience. we use kvm and virt-manager, but sometimes i'm not really sure about the whether it's kvm or virt-manager problem so i collect them together. our system: MB: Intel S3000AHV CPU: Intel Core 2 Quad CPU Q6600 @ 2.40GHz RAM: 8GB CentOS 5 x86-64 host system with: kernel-2.6.18-8.1.8.el5 kmod-kvm-35-1.2.6.18_8.1.8.el5 kvm-35-1 libvirt-0.3.2-2 libvirt-python-0.3.2-2 python-virtinst-0.300.0-1 virt-manager-0.5.0-1 virt-top-0.3.2.5-1 virt-viewer-0.0.2-1 kvm is not ready for production use for many reason: it can't reboot which imho a very basic feature, what's more can't even shutdown/poweroff. the centos i386 guest are not able to shutdown on the x86_64 host (strange the x86_64 centos and the i586 mandrake-9 are able to shutdown) or what's seems to be more likely it's a random think. can't give 2GB ram to any guests not even 2, but 1.9999 is ok. what's more i've to discover this in the hard way virt-manager gives some strange error message (anyway none of the virt-manager's error message are useful, just strange stack trace without any kind of useful info). the host see as i've 4 cpu. i've got a change to gives more cpu to the guest, what's more they starts, but after a few minutes running the system crash. not just the guest os but the host os crash without any kind of info, log or any useful info what was the cause of it, but hard reset helps:-((( first of all it's serious problem, even if it's a known bug and documented somewhere (but there is not any kind of docs about neither kvm nor virt-manager/libvirt and what i can find that very limited), since all of the new hvm cpu (which is required for kvm) has more core, so the guest can't use the real power of the cpu this means i've to put a lots of guests with one virtual cpu to the host or currently can't exploit these cpus. libvirt not start the guest when the host started, not saved the state of the guests when stopped, when restarted all guest get into shutdown state. may be it's just libvirt but, but if kvm not able to save it's guest's state that it's a real pain. this behavior not acceptable in a real environment. in virt-manager can't modify virtual cpu number of the guests (virt-manager crash with another unusable stacks trace). it i modify the xml config file by hand i've to restart libvirt in order to read it. even if i just modify one guest's config, but as if i restart libvirtd the all other guest will be killed it's another nice feature. from vit-manager i can't assign swap partitions (and any other device to the guests) during guest creation. i precreate lvm partition on the host for root and swap to all of our guests (about 10) but it's a real pain to reload libvirtd (which kills all guests) to add a new disk to the guest which can be used as a swap or other partition. what's more it'd be useful to add swap partition during install. the same thing apply to cdrom too. if i made a mistake in the libvirtd xml config file, then i don't get any kind of error message just libvirt don't start the given guest. no error message service seems to start without error and don't know what was the reason that the guest not start (sometimes because syntactic error in the config file, sometimes kvm/qemu not support it eg more then 2gb mem). there are unwritten unix/linux tradition eg if there are error during a service startup report it some log file. if a service can't start without error then it not start at all eg. if i've apache with 5 virtual host but one of it's config has error then the whole httpd not started. it'd have to be the same with libvirt too. libvirt config file has all the features as qemu command line (if not it can't be used by advanced users)? but i don't know since there is not a reference docs for the config file. there is not a detailed reference docs about libvirtd config (and at the same time the gui not contains all the settings). for the the xml format is a bit redundant eg. "source dev", "target dev" why not just source and target, inside interface type='bridge' 'source bridge' why not just source (i've to modify two place when i'd like to switch from network to bride. ok these are just small cosmetic thing just mention. when create a new guest in virt-manager and reach the last page and press finish then most of the time (2 times from 3) i've got an error something like "can't connect to qemu" after i press ok and press again finish then it's able to create the new host (so always have to retry it). in virt-manager when a guest is shutdown and double-click on the guest and try to start it it's not possible. you've to select the guest in the main windows, right-click and start it. the new vnc based console is very unstable or just virt-manager not able to connect it. many times i double click on a guest in virt-manager i got the window can't connect to console while the guest are running. that's why i like to see again the serial console. we've got other virtual servers with xen and openvz, with has other problems but works. my conclusion is that currently kvm, libvirtd, virt-manager (et-mgmt-tools) seems very promising. but currently just in a very alpha stage using even in non mission critical but production environment still a very 'brave' decision. xen has a few years vantage, but it's at least usable in real environment. i'm on the these list and see what kind of problem you're working which is great but at the same it'd be useful to fix the most basic problems with these tools. just my 2c to these community. -- Levente "Si vis pacem para bellum!" From lfarkas at bppiac.hu Sat Sep 8 13:06:31 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Sat, 08 Sep 2007 15:06:31 +0200 Subject: [et-mgmt-tools] Re: [kvm-devel] (long) kvm and virt-manager not ready for daily usage In-Reply-To: <68676e00709080535ie3f6e56l822041e307cc16e1@mail.gmail.com> References: <46E28438.5000403@bppiac.hu> <68676e00709080535ie3f6e56l822041e307cc16e1@mail.gmail.com> Message-ID: <46E29E57.7070104@bppiac.hu> Luca wrote: > On 9/8/07, Farkas Levente wrote: >> kvm is not ready for production use for many reason: >> it can't reboot which imho a very basic feature, what's more can't even >> shutdown/poweroff. the centos i386 guest are not able to shutdown on the >> x86_64 host (strange the x86_64 centos and the i586 mandrake-9 are able >> to shutdown) or what's seems to be more likely it's a random think. > > Reboot failure is a regression; for shutdown you should enable APCI > with "acpi=force"; bochs BIOS doesn't have SMBIOS/DMI tables and Linux > may refuse to use ACPI without them (it depends on the configuration). when, where, how? is there any docs about it? -- Levente "Si vis pacem para bellum!" From lfarkas at bppiac.hu Sat Sep 8 20:50:02 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Sat, 08 Sep 2007 22:50:02 +0200 Subject: [et-mgmt-tools] Re: [kvm-devel] (long) kvm and virt-manager not ready for daily usage In-Reply-To: <1189264729.19248.3.camel@squirrel> References: <46E28438.5000403@bppiac.hu> <1189264729.19248.3.camel@squirrel> Message-ID: <46E30AFA.5060800@bppiac.hu> Anthony Liguori wrote: >> the host see as i've 4 cpu. i've got a change to gives more cpu to the >> guest, what's more they starts, but after a few minutes running the >> system crash. not just the guest os but the host os crash without any >> kind of info, log or any useful info what was the cause of it, but hard >> reset helps:-((( first of all it's serious problem, even if it's a known >> bug and documented somewhere (but there is not any kind of docs about >> neither kvm nor virt-manager/libvirt and what i can find that very >> limited), since all of the new hvm cpu (which is required for kvm) has >> more core, so the guest can't use the real power of the cpu this means >> i've to put a lots of guests with one virtual cpu to the host or >> currently can't exploit these cpus. > > I'm having difficulty understanding what your problem is. Are you after i reread my sentences it was difficult for me too:-( > saying that guest SMP isn't working for you? The host OS definitely > shouldn't crash. Can you be more specific about what configs you are > using? There was a host oops fixed in kvm-36 so upgrading may help you. exactly. i've got 4 phisical core (Intel Core 2 Quad) and i try to give 4 cpus for 2 guests and 2 cpus for a third guest and restart libvirtd. the result was that even the host system crash without any stack trace or kernel panic and only the hard reset helps. -- Levente "Si vis pacem para bellum!" From mdehaan at redhat.com Mon Sep 10 14:22:59 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 10 Sep 2007 10:22:59 -0400 Subject: [et-mgmt-tools] (long) kvm and virt-manager not ready for daily usage In-Reply-To: <46E28438.5000403@bppiac.hu> References: <46E28438.5000403@bppiac.hu> Message-ID: <46E55343.6070804@redhat.com> Farkas Levente wrote: [snip] Please do not cross-post to this list. It makes mail filtering with threading intact very difficult. --Michael From kevin at intriguedsolutions.com Mon Sep 10 17:33:25 2007 From: kevin at intriguedsolutions.com (kevin at intriguedsolutions.com) Date: Mon, 10 Sep 2007 13:33:25 -0400 Subject: [et-mgmt-tools] Local cobbler install for closed cctv network Message-ID: <20070910133325.we3ebcps00440w8k@69.89.31.112> Anyone have an idea why I cannot install cobbler 0.5.1 or 0.6.1 on a Fedora Core 6 install? I keep getting errors stating no network access, cannot find baseurl, and no packages were given. I tried editing out lines in the .repo files but doesn't sound like a thing to do. However, my job is closed network and I need to get an image server up fast. Our Red Hat 8 server is not sending the images anymore. I am new to Linux but I can catch on quick if I had some step by step instructions. Am I missing something? Thanks From wright at imageworks.com Mon Sep 10 17:36:47 2007 From: wright at imageworks.com (Peter Wright) Date: Mon, 10 Sep 2007 10:36:47 -0700 Subject: [et-mgmt-tools] Local cobbler install for closed cctv network In-Reply-To: <20070910133325.we3ebcps00440w8k@69.89.31.112> References: <20070910133325.we3ebcps00440w8k@69.89.31.112> Message-ID: <46E580AF.3010605@imageworks.com> kevin at intriguedsolutions.com wrote: > > Anyone have an idea why I cannot install cobbler 0.5.1 or 0.6.1 on a > Fedora Core 6 install? I keep getting errors stating no network > access, cannot find baseurl, and no packages were given. I tried > editing out lines in the .repo files but doesn't sound like a thing to > do. However, my job is closed network and I need to get an image > server up fast. Our Red Hat 8 server is not sending the images > anymore. I am new to Linux but I can catch on quick if I had some step > by step instructions. Am I missing something? > > It is pretty dificult to help out unless you post the error messages you are getting, as well as specifics about your current configuration. Taking a wild guess, I bet you are using the default FC6 /etc/yum.repos.d/* files which attempt to contact a yum repository that is on the public internet. Although, this is just a guess. At what point are you running into problems? How far have you gotten into your FC6 install? Do you need to use something like cobbler, or can you just create a kickstart and start the installs by hand? -pete -- Peter Wright Systems Administrator Sony Pictures Imageworks wright at imageworks.com www.imageworks.com From mdehaan at redhat.com Mon Sep 10 18:44:53 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 10 Sep 2007 14:44:53 -0400 Subject: [et-mgmt-tools] Local cobbler install for closed cctv network In-Reply-To: <20070910133325.we3ebcps00440w8k@69.89.31.112> References: <20070910133325.we3ebcps00440w8k@69.89.31.112> Message-ID: <46E590A5.5080403@redhat.com> kevin at intriguedsolutions.com wrote: > Anyone have an idea why I cannot install cobbler 0.5.1 or 0.6.1 on a > Fedora Core 6 install? I keep getting errors stating no network > access, cannot find baseurl, and no packages were given. This really isn't enough information to go on. I'd need a rundown on the steps you performed, what you have in your settings file, and so on. The easiest way to get through this is probably to join #cobbler on irc.freenode.net, where I or some other folks can help debug your setup. > I tried editing out lines in the .repo files but doesn't sound like a > thing to do. However, my job is closed network and I need to get an > image server up fast. Our Red Hat 8 server is not sending the images > anymore. I am new to Linux but I can catch on quick if I had some step > by step instructions. Am I missing something? Sounds like you've mirrored the wrong URLs to me, or have ran "cobbler repo add" commands without the associated "cobbler reposync" to pull content down. > > 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 Tue Sep 11 01:18:19 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Tue, 11 Sep 2007 02:18:19 +0100 Subject: [et-mgmt-tools] FYI: virt-manager improved remote management Message-ID: <20070911011819.GC12579@redhat.com> Having played around with the remote management in virt-manager 0.5.0 with a little more seriousness there were some fairly obvious issues to fix. eg If the remote host is hung / slow to respond the initial connection attempt will hang the whole UI. Also if SSH keys & agent were not setup then when using a connection tunnelled over SSH the whole app could sit there forever waiting for the user to enter a password on the command line - which you'd never see in a GUI app unless you happened to launch from a terminal. So I've refactored things a little. First of all virt-manager will fork into the background, close STDIN and setsid() to drop its controlling TTY. This prevents any attempts by SSH to prompt for terminal input. Second, I've moved the calls to libvirt.open() into a background thread. While virt-manager is attempting to make a connection you'll see a 'Connecting' status against the row. You'll get a notification if there's a failure as before. You can have many connections all attempting to conect at once if you want.. 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 lfarkas at bppiac.hu Tue Sep 11 10:21:40 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Tue, 11 Sep 2007 12:21:40 +0200 Subject: [et-mgmt-tools] latest virt-manager worse then ever Message-ID: <46E66C34.90700@bppiac.hu> hi, with latest virt-manager and libvirt libvirt-0.3.2-2 libvirt-python-0.3.2-2 python-virtinst-0.300.0-1 virt-manager-0.5.0-1 we're not bale to install any new guests:-( when try to create a new guest and press finish for the first time we've got this error: ---------------------------------------- Unable to complete install 'libvirt.libvirtError virDomainCreateLinux() failed Timed out while reading monitor startup output Traceback (most recent call last): File "/usr/share/virt-manager/virtManager/create.py", line 612, in do_install dom = guest.start_install(False, meter = meter) File "/usr/lib/python2.4/site-packages/virtinst/Guest.py", line 718, in start_install return self._do_install(consolecb, meter) File "/usr/lib/python2.4/site-packages/virtinst/Guest.py", line 735, in _do_install self.domain = self.conn.createLinux(install_xml, 0) File "/usr/lib64/python2.4/site-packages/libvirt.py", line 563, in createLinux if ret is None:raise libvirtError('virDomainCreateLinux() failed', conn=self) libvirtError: virDomainCreateLinux() failed Timed out while reading monitor startup output ' ---------------------------------------- and if we press finish again and again we lost cdrom: ---------------------------------------- Unable to complete install 'exceptions.ValueError The CDROM device hdc is already used Traceback (most recent call last): File "/usr/share/virt-manager/virtManager/create.py", line 612, in do_install dom = guest.start_install(False, meter = meter) File "/usr/lib/python2.4/site-packages/virtinst/Guest.py", line 718, in start_install return self._do_install(consolecb, meter) File "/usr/lib/python2.4/site-packages/virtinst/Guest.py", line 731, in _do_install install_xml = self.get_config_xml() File "/usr/lib/python2.4/site-packages/virtinst/Guest.py", line 683, in get_config_xml return """ File "/usr/lib/python2.4/site-packages/virtinst/FullVirtGuest.py", line 187, in _get_device_xml return """ File "/usr/lib/python2.4/site-packages/virtinst/Guest.py", line 661, in _get_device_xml return """%(disks)s File "/usr/lib/python2.4/site-packages/virtinst/FullVirtGuest.py", line 254, in _get_disk_xml raise ValueError, "The CDROM device %s is already used" % d.target ValueError: The CDROM device hdc is already used ' ---------------------------------------- so currently it's not possible to create any new guest:-( -- Levente "Si vis pacem para bellum!" From rjones at redhat.com Tue Sep 11 10:58:13 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Tue, 11 Sep 2007 11:58:13 +0100 Subject: [et-mgmt-tools] latest virt-manager worse then ever In-Reply-To: <46E66C34.90700@bppiac.hu> References: <46E66C34.90700@bppiac.hu> Message-ID: <46E674C5.1050507@redhat.com> Dear Farkas, Do you think that posting "latest virt-manager worse then ever" is actually helping anything, or just annoying and disheartening for all the people who are working very hard to make virt-manager, KVM and the other tools better? Constructive criticism is fine, and I agree with you that virt-manager isn't as smooth as it could be. This is mainly caused by the inherent difficulty of bringing together so many different and incompatible systems together to do what is a very complex and multi-faceted job. Also, looking up bugs in Bugzilla and reporting new ones is helpful. So is contributing patches and fixes. 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 Hua.Zhang at Sun.COM Tue Sep 11 13:22:45 2007 From: Hua.Zhang at Sun.COM (Henry Zhang) Date: Tue, 11 Sep 2007 21:22:45 +0800 Subject: [et-mgmt-tools] Bug in serial console Message-ID: <46E696A5.3000304@sun.com> Hi there, I am trying to run virt-manager at Solaris, now seems all is fine, but after I restore or run a domain which is saved, shutdown or paused, I find I can't find any prompt in serial console any more, and after I exited virt-manger, and run it again, I can see the prompt in serial console, seems serial console does not clean up correctly or didn't connect itself correct? I am using virt-manger 0.4.0. Not sure if there is a similar problem in Linux? And is it fixed? Thanks, Henry From mdehaan at redhat.com Tue Sep 11 21:36:47 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 11 Sep 2007 17:36:47 -0400 Subject: [et-mgmt-tools] Cobbler XMLRPC API (for 0.6.2) and pending Web UI Message-ID: <46E70A6F.9030200@redhat.com> I've checked in a first-pass at what will be the new Cobbler "Read/Write" XMLRPC API, which is important of the development of the Web UI and letting other 3rd party apps interface with Cobbler. Prior to this, it was possible to query cobbler remotely, but it was not possible to change anything. This will be available at http(s)://server.example.org/cobbler_api_rw and disabled by default. (Yes -- if you've not heard, someone is developing a Cobbler Web UI -- major thanks go out to Al Tobey -- ... we'll have it checked in to git shortly so more folks can look at it, build it as part of the RPM, and help out --- it's all Python and uses Cheetah templates, just like Cobbler) There is some documentation on the new API here: https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerXmlrpc Once the Web UI is checked in and packaged, I'll send out another email about how to start playing with it (at least in a display-only sort of mode as the other features get added). It will probably get included in some form in the next release and upgraded in subsequent releases. --Michael From kevin at intriguedsolutions.com Wed Sep 12 00:25:32 2007 From: kevin at intriguedsolutions.com (Kevin Pescatello) Date: Tue, 11 Sep 2007 20:25:32 -0400 Subject: [et-mgmt-tools] FYI: virt-manager improved remote management In-Reply-To: <20070911011819.GC12579@redhat.com> References: <20070911011819.GC12579@redhat.com> Message-ID: <000f01c7f4d3$6f02f4d0$4d08de70$@com> Thanks for the help. I figured out Red Hat install and we have our image-server back. KP -----Original Message----- From: Daniel P. Berrange [mailto:berrange at redhat.com] Sent: Monday, September 10, 2007 9:18 PM To: et-mgmt-tools at redhat.com Subject: [et-mgmt-tools] FYI: virt-manager improved remote management Having played around with the remote management in virt-manager 0.5.0 with a little more seriousness there were some fairly obvious issues to fix. eg If the remote host is hung / slow to respond the initial connection attempt will hang the whole UI. Also if SSH keys & agent were not setup then when using a connection tunnelled over SSH the whole app could sit there forever waiting for the user to enter a password on the command line - which you'd never see in a GUI app unless you happened to launch from a terminal. So I've refactored things a little. First of all virt-manager will fork into the background, close STDIN and setsid() to drop its controlling TTY. This prevents any attempts by SSH to prompt for terminal input. Second, I've moved the calls to libvirt.open() into a background thread. While virt-manager is attempting to make a connection you'll see a 'Connecting' status against the row. You'll get a notification if there's a failure as before. You can have many connections all attempting to conect at once if you want.. Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| _______________________________________________ et-mgmt-tools mailing list et-mgmt-tools at redhat.com https://www.redhat.com/mailman/listinfo/et-mgmt-tools From berrange at redhat.com Wed Sep 12 02:54:50 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 12 Sep 2007 03:54:50 +0100 Subject: [et-mgmt-tools] virt-install PXE boot support Message-ID: <20070912025450.GD31998@redhat.com> I have just pushed a change to virt-install which allows PXE boot installs to be done for fullyvirtualized guests. Instead of using the --location or --cdrom arggs to virt-install, use the cunningly named --pxe argument NB, this currently only works with Xen fullyvirt due to a bug in the libvirt code for parsing KVM pxe XML configs. I will fix this shortly. 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 mike at mnmoran.org Thu Sep 13 11:56:39 2007 From: mike at mnmoran.org (Michael N. Moran) Date: Thu, 13 Sep 2007 07:56:39 -0400 Subject: [et-mgmt-tools] Configuring a cdrom Message-ID: <46E92577.2040201@mnmoran.org> Using my shinny new notebook with an Intel T7300 (VT) and FC7 installed, I successfully created a virtual machine running Windoze XP SP2 using virt-manager under KVM. The only hiccup came after the first reboot when I lost my CD-ROM "device". Googling and fiddling led me to delete-and-then "re-create" using the semi-installed image. This allowed me to finish the install. Networking and *many* Windoze things work! I *love* seeing Windoze-in-a-window :-) However, I am now unable to find a way to attach a CD-ROM to install *other* software. I've googled and searched for config files to no avail. The XML config file in /etc/libvirt/qemu seems to be created by libvirt/virt-manager, but not read by the virt-manager, since I can delete it and re-launch virt-manager and *still* have my virtual machine configuration ... I tried using virt-manager itself, but cannot seem to find any graphical way to add a CD-ROM. Please enlighten me that my joy may be complete! Here's my config: Fedora Core 7 Dell Latitude D830 Linux boogie 2.6.22.4-65.fc7 #1 SMP Tue Aug 21 21:50:50 EDT 2007 x86_64 x86_64 x86_64 GNU/Linux virt-manager-0.4.0-2.fc7 libvirt-0.3.2-1.fc7 kvm-24-1 -- Michael N. Moran (h) 770 516 7918 5009 Old Field Ct. (c) 678 521 5460 Kennesaw, GA, USA 30144 http://mnmoran.org "So often times it happens, that we live our lives in chains and we never even know we have the key." "Already Gone" by Jack Tempchin (recorded by The Eagles) The Beatles were wrong: 1 & 1 & 1 is 1 From berrange at redhat.com Thu Sep 13 13:02:13 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 13 Sep 2007 14:02:13 +0100 Subject: [et-mgmt-tools] Configuring a cdrom In-Reply-To: <46E92577.2040201@mnmoran.org> References: <46E92577.2040201@mnmoran.org> Message-ID: <20070913130213.GA23525@redhat.com> On Thu, Sep 13, 2007 at 07:56:39AM -0400, Michael N. Moran wrote: > > > Using my shinny new notebook with an Intel T7300 (VT) > and FC7 installed, I successfully created a virtual > machine running Windoze XP SP2 using virt-manager > under KVM. > > The only hiccup came after the first reboot when I > lost my CD-ROM "device". Googling and fiddling led > me to delete-and-then "re-create" using the semi-installed > image. This allowed me to finish the install. > > Networking and *many* Windoze things work! I *love* > seeing Windoze-in-a-window :-) > > However, I am now unable to find a way to attach > a CD-ROM to install *other* software. > > I've googled and searched for config files to no > avail. The XML config file in /etc/libvirt/qemu > seems to be created by libvirt/virt-manager, but > not read by the virt-manager, since I can delete > it and re-launch virt-manager and *still* have my > virtual machine configuration ... The config files are read by the libvirt daemon, so if you change them you need to do '/etc/init.d/libvirtd reload' to make it detect changes. > I tried using virt-manager itself, but cannot seem > to find any graphical way to add a CD-ROM. The version of virt-manager currently in F7 does not have ability to add a CDROM device - it can only add regular disks. We added support for CDROM in 0.5.0, but that release is too buggy to push to F7 at this time. So at this time either use a disk, or edit the config file in /etc/libvirt to add the CDROM 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 mike at mnmoran.org Thu Sep 13 13:58:45 2007 From: mike at mnmoran.org (Michael N. Moran) Date: Thu, 13 Sep 2007 09:58:45 -0400 Subject: [et-mgmt-tools] Configuring a cdrom In-Reply-To: <20070913130213.GA23525@redhat.com> References: <46E92577.2040201@mnmoran.org> <20070913130213.GA23525@redhat.com> Message-ID: <46E94215.1040402@mnmoran.org> Daniel P. Berrange wrote: > On Thu, Sep 13, 2007 at 07:56:39AM -0400, Michael N. Moran wrote: >> I've googled and searched for config files to no >> avail. The XML config file in /etc/libvirt/qemu >> seems to be created by libvirt/virt-manager, but >> not read by the virt-manager, since I can delete >> it and re-launch virt-manager and *still* have my >> virtual machine configuration ... > > The config files are read by the libvirt daemon, so if you change them > you need to do '/etc/init.d/libvirtd reload' to make it detect changes. That's it! I love you guys ;-) I should have known that the libvirtd needed a kick in the pants.... how many years have I been running Linux? Now I can use Visio for my UML diagrams at work and Linux for everything else! Now if I can just get open-office to do a better UML drawing package ... thanks Dan -- Michael N. Moran (h) 770 516 7918 5009 Old Field Ct. (c) 678 521 5460 Kennesaw, GA, USA 30144 http://mnmoran.org "So often times it happens, that we live our lives in chains and we never even know we have the key." "Already Gone" by Jack Tempchin (recorded by The Eagles) The Beatles were wrong: 1 & 1 & 1 is 1 From berrange at redhat.com Thu Sep 13 14:42:18 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 13 Sep 2007 15:42:18 +0100 Subject: [et-mgmt-tools] virt-install PXE boot support In-Reply-To: <20070912025450.GD31998@redhat.com> References: <20070912025450.GD31998@redhat.com> Message-ID: <20070913144218.GE23525@redhat.com> On Wed, Sep 12, 2007 at 03:54:50AM +0100, Daniel P. Berrange wrote: > I have just pushed a change to virt-install which allows PXE boot installs > to be done for fullyvirtualized guests. Instead of using the --location > or --cdrom arggs to virt-install, use the cunningly named --pxe argument I have now also added UI in virt-manager to allow fullyvirtualized guest to be PXE booted. This depends on the newest code from virt-install HG repo. 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 Thu Sep 13 21:52:44 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 13 Sep 2007 22:52:44 +0100 Subject: [et-mgmt-tools] [FW: [gtk-vnc-devel] ANNOUNCE: Release 0.2.0 of GTK-VNC] Message-ID: <20070913215244.GB2046@redhat.com> ----- Forwarded message from "Daniel P. Berrange" ----- > Date: Thu, 13 Sep 2007 22:49:16 +0100 > From: "Daniel P. Berrange" > To: gtk-vnc-devel > Subject: [gtk-vnc-devel] ANNOUNCE: Release 0.2.0 of GTK-VNC > > The GTK-VNC development team is pleased to announce the availability of > release 0.2.0 > > It can be downloaded from sourceforge at: > > http://downloads.sourceforge.net/gtk-vnc/gtk-vnc-0.2.0.tar.gz > > Bug fixes: > > This release fixes compatability with mixed endianess client > and servers and BGR/RGB pixel format displays. Mixed colour > depths are now correctly accounted for & converted appropriately. > Graceful shutdown is now handled correctly, interrupting any > pending I/O waits. Added destructor to widget to ensure all > resources are free'd when the widget is deleted. > > New features: > > The two common client cursor offload extensions are supported > for compatability with OSX-VNC, UltraVNC and TightVNC. Support > for Vino's TLS authentication method is added. Added signals > for getting notified of desktop resize events, and authentication > failures. > > Further information is available from the project website at > > http://gtk-vnc.sourceforge.net/ > > Regards, > Dan (on behalf of the GTK-VNC development team) ----- End forwarded message ----- 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 Fri Sep 14 17:07:41 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Fri, 14 Sep 2007 18:07:41 +0100 Subject: [et-mgmt-tools] Experimental virt-p2v (physical to virtual server migration) tool Message-ID: <46EABFDD.6070105@redhat.com> I've been writing an experimental P2V (physical to virtual server migration) tool. It's a live CD, the idea is that you boot it on the physical server and it examines the local disks and ships them off to your Xen/whatever host where they can be turned easily into a replica virtual machine. The concept works, but the current implementation is a bit broken. I hope to have something working a bit better over the weekend. http://et.redhat.com/~rjones/virt-p2v/ Nevertheless, it's just a big shell script so if anyone feels like lending a hand then send me an email, or ping me on #virt (@ OFTC). 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 rjones at redhat.com Sat Sep 15 13:35:54 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Sat, 15 Sep 2007 14:35:54 +0100 Subject: [et-mgmt-tools] Experimental virt-p2v (physical to virtual server migration) tool In-Reply-To: <46EABFDD.6070105@redhat.com> References: <46EABFDD.6070105@redhat.com> Message-ID: <46EBDFBA.30400@redhat.com> Richard W.M. Jones wrote: > I've been writing an experimental P2V (physical to virtual server > migration) tool. It's a live CD, the idea is that you boot it on the > physical server and it examines the local disks and ships them off to > your Xen/whatever host where they can be turned easily into a replica > virtual machine. > > The concept works, but the current implementation is a bit broken. I > hope to have something working a bit better over the weekend. > > http://et.redhat.com/~rjones/virt-p2v/ > > Nevertheless, it's just a big shell script so if anyone feels like > lending a hand then send me an email, or ping me on #virt (@ OFTC). Check out version 0.4 from the above site. It __almost__ works; at least the TCP transport works, but the ssh transport for some reason fails to be able to connect to the remote machine (even though I can 'ssh' perfectly fine from another virtual console). Again, all feedback, patches, etc. gratefully received. (Also I've clarified that the license is GPL). 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 jensampol at yahoo.de Sun Sep 16 23:21:53 2007 From: jensampol at yahoo.de (Jens Ahrens) Date: Mon, 17 Sep 2007 01:21:53 +0200 (CEST) Subject: [et-mgmt-tools] cobbler-koan: installer asks for ip, netmask, gateway, dns Message-ID: <853795.6579.qm@web25506.mail.ukl.yahoo.com> I'm running FC7 with kernel-xen-2.6.20-2931.fc7 xen-libs-3.1.0-2.fc7 xen-3.1.0-2.fc7 cobbler-0.6.1-2.fc7 I'm trying to install a paravirtual machine (Scientific Linux 4.5) with a fixed ip-address(no private ip) and hostname and a kickstartfile accessible via e.g. http (nfs would also be possible). Everything seems to work, but the installation tries to determine the IP for eth0, which ends up in a menu asking for ip, netmask, gateway and dns. If I do the input manually the installtion continue using the values provided by my kickstart file. How can I pass the values like ip, netmask, gateway and dns to the SL4.5 installer, without having to type them in manually? Thanks in advance Wissenswertes f?r Bastler und Hobby Handwerker. BE A BETTER HEIMWERKER! www.yahoo.de/clever From thestrider at gmail.com Mon Sep 17 00:18:08 2007 From: thestrider at gmail.com (Adam Rosenwald) Date: Sun, 16 Sep 2007 20:18:08 -0400 Subject: [et-mgmt-tools] cobbler-koan: installer asks for ip, netmask, gateway, dns In-Reply-To: <853795.6579.qm@web25506.mail.ukl.yahoo.com> References: <853795.6579.qm@web25506.mail.ukl.yahoo.com> Message-ID: <46EDC7C0.6020506@gmail.com> I too have issues automating static network configuration within anaconda. I'm using a RHEL5 domU kickstart with the following standard static network assignment stanza: network --bootproto=static --device=eth0 --ip= --gateway= --netmask= --hostname= --noipv6 --onboot=yes I've also set --kopts='noipv6 nodhcp'. While this works fine when doing a standard (non-virtual) install, I receive the same menu prompt that Jens encounters. More specifically, IPV4 -- checked on. DHCP is checked on. Manual is checked off IPV6 -- checked off. This leads me to believe that there must be additional kernel options not mentioned in the documentation (http://www.redhat.com/docs/manuals/enterprise/RHEL-5-manual/Installation_Guide-en-US/ch-bootopts-x86.html). Strictly speaking, our question belongs on the kickstart mailing list; however, this whole process falls within the context of provisioning virtual systems within cobbler. Thx, -A. Jens Ahrens wrote: > I'm running FC7 with > kernel-xen-2.6.20-2931.fc7 > xen-libs-3.1.0-2.fc7 > xen-3.1.0-2.fc7 > cobbler-0.6.1-2.fc7 > > I'm trying to install a paravirtual machine > (Scientific Linux 4.5) with a fixed ip-address(no > private ip) and hostname and a kickstartfile > accessible via e.g. http (nfs would also be possible). > Everything seems to work, but the installation tries > to determine the IP for eth0, which ends up in a menu > asking for ip, netmask, gateway and dns. > If I do the input manually the installtion continue > using the values provided by my kickstart file. > > How can I pass the values like ip, netmask, gateway > and dns to the SL4.5 installer, without having to type > them in manually? > > Thanks in advance > > > > Wissenswertes f?r Bastler und Hobby Handwerker. BE A BETTER HEIMWERKER! www.yahoo.de/clever > > _______________________________________________ > 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 Sep 17 14:43:49 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 17 Sep 2007 10:43:49 -0400 Subject: [et-mgmt-tools] cobbler-koan: installer asks for ip, netmask, gateway, dns In-Reply-To: <46EDC7C0.6020506@gmail.com> References: <853795.6579.qm@web25506.mail.ukl.yahoo.com> <46EDC7C0.6020506@gmail.com> Message-ID: <46EE92A5.1030602@redhat.com> Adam Rosenwald wrote: > Strictly speaking, our question belongs on the kickstart mailing list; > however, this whole process falls within the context of provisioning > virtual systems within cobbler. Thx, > > -A. > Yes, I'd take this up with kickstart-list if you are having problems configuring static IP's. You said: "This leads me to believe that there must be additional kernel options not mentioned in the documentation " I really doubt it, though it's hard to tell what's going on just from your email. I would take this up with kickstart-list at redhat.com as networking setups do come up there quite a lot. I've also created a Google search engine for it here that turns up a lot of threads on the subject of static IP configuration ... http://google.com/coop/cse?cx=016811804524159694721%3A1h7btspnxtu If you can't find an answer, you could also boot DHCP and just use the kickstart to configure the static configuration, or you could also use DHCP reservations for IP assignment (which cobbler makes quite easy to handle with manage_dhcp enabled). > Jens Ahrens wrote: > >> I'm running FC7 with >> kernel-xen-2.6.20-2931.fc7 >> xen-libs-3.1.0-2.fc7 >> xen-3.1.0-2.fc7 >> cobbler-0.6.1-2.fc7 >> >> I'm trying to install a paravirtual machine >> (Scientific Linux 4.5) with a fixed ip-address(no >> private ip) and hostname and a kickstartfile >> accessible via e.g. http (nfs would also be possible). >> Everything seems to work, but the installation tries >> to determine the IP for eth0, which ends up in a menu >> asking for ip, netmask, gateway and dns. >> If I do the input manually the installtion continue >> using the values provided by my kickstart file. >> >> How can I pass the values like ip, netmask, gateway >> and dns to the SL4.5 installer, without having to type >> them in manually? >> >> Thanks in advance >> >> >> >> Wissenswertes f?r Bastler und Hobby Handwerker. BE A BETTER HEIMWERKER! www.yahoo.de/clever >> >> _______________________________________________ >> 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 thestrider at gmail.com Mon Sep 17 17:55:25 2007 From: thestrider at gmail.com (Adam Rosenwald) Date: Mon, 17 Sep 2007 13:55:25 -0400 Subject: [et-mgmt-tools] cobbler-koan: installer asks for ip, netmask, gateway, dns In-Reply-To: <46EE92A5.1030602@redhat.com> References: <853795.6579.qm@web25506.mail.ukl.yahoo.com> <46EDC7C0.6020506@gmail.com> <46EE92A5.1030602@redhat.com> Message-ID: <658267190709171055n263cd279n1db333c031e5d0b3@mail.gmail.com> Thanks. Moving to kickstart-list. On 9/17/07, Michael DeHaan wrote: > > Adam Rosenwald wrote: > > Strictly speaking, our question belongs on the kickstart mailing list; > > however, this whole process falls within the context of provisioning > > virtual systems within cobbler. Thx, > > > > -A. > > > > Yes, I'd take this up with kickstart-list if you are having problems > configuring static IP's. > > You said: "This leads me to believe that there must be additional > kernel options not mentioned in the documentation " > > I really doubt it, though it's hard to tell what's going on just from > your email. I would take this up with kickstart-list at redhat.com as > networking setups do come up there quite a lot. I've also created a > Google search engine for it here that turns up a lot of threads on the > subject > of static IP configuration ... > http://google.com/coop/cse?cx=016811804524159694721%3A1h7btspnxtu > > If you can't find an answer, you could also boot DHCP and just use the > kickstart to configure the static configuration, or you could > also use DHCP reservations for IP assignment (which cobbler makes quite > easy to handle with manage_dhcp enabled). > > > > Jens Ahrens wrote: > > > >> I'm running FC7 with > >> kernel-xen-2.6.20-2931.fc7 > >> xen-libs-3.1.0-2.fc7 > >> xen-3.1.0-2.fc7 > >> cobbler-0.6.1-2.fc7 > >> > >> I'm trying to install a paravirtual machine > >> (Scientific Linux 4.5) with a fixed ip-address(no > >> private ip) and hostname and a kickstartfile > >> accessible via e.g. http (nfs would also be possible). > >> Everything seems to work, but the installation tries > >> to determine the IP for eth0, which ends up in a menu > >> asking for ip, netmask, gateway and dns. > >> If I do the input manually the installtion continue > >> using the values provided by my kickstart file. > >> > >> How can I pass the values like ip, netmask, gateway > >> and dns to the SL4.5 installer, without having to type > >> them in manually? > >> > >> Thanks in advance > >> > >> > >> > >> Wissenswertes f?r Bastler und Hobby Handwerker. BE A BETTER > HEIMWERKER! www.yahoo.de/clever > >> > >> _______________________________________________ > >> 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 > -------------- next part -------------- An HTML attachment was scrubbed... URL: From dlutter at redhat.com Mon Sep 17 17:56:01 2007 From: dlutter at redhat.com (David Lutterkort) Date: Mon, 17 Sep 2007 17:56:01 +0000 Subject: [et-mgmt-tools] Can't make the meeting tomorrow Message-ID: <1190051761.20285.2.camel@localhost.localdomain> I will be at IDF, having lots of fun. David From thestrider at gmail.com Mon Sep 17 18:13:53 2007 From: thestrider at gmail.com (Adam Rosenwald) Date: Mon, 17 Sep 2007 14:13:53 -0400 Subject: [et-mgmt-tools] cobbler-koan: installer asks for ip, netmask, gateway, dns In-Reply-To: <658267190709171055n263cd279n1db333c031e5d0b3@mail.gmail.com> References: <853795.6579.qm@web25506.mail.ukl.yahoo.com> <46EDC7C0.6020506@gmail.com> <46EE92A5.1030602@redhat.com> <658267190709171055n263cd279n1db333c031e5d0b3@mail.gmail.com> Message-ID: <658267190709171113p688e54eei4db50e6dec9ae856@mail.gmail.com> Jens et. al, Please refer to https://www.redhat.com/archives/kickstart-list/2007-August/msg00068.html I've confirmed that this works for virtual deployments via cobbler/koan. Note that this is NOT located in the aforementioned Red Hat documentation -- but should be. -A. On 9/17/07, Adam Rosenwald wrote: > > Thanks. Moving to kickstart-list. > > On 9/17/07, Michael DeHaan wrote: > > > > Adam Rosenwald wrote: > > > Strictly speaking, our question belongs on the kickstart mailing list; > > > however, this whole process falls within the context of provisioning > > > virtual systems within cobbler. Thx, > > > > > > -A. > > > > > > > Yes, I'd take this up with kickstart-list if you are having problems > > configuring static IP's. > > > > You said: "This leads me to believe that there must be additional > > kernel options not mentioned in the documentation " > > > > I really doubt it, though it's hard to tell what's going on just from > > your email. I would take this up with kickstart-list at redhat.com as > > networking setups do come up there quite a lot. I've also created a > > Google search engine for it here that turns up a lot of threads on the > > subject > > of static IP configuration ... > > http://google.com/coop/cse?cx=016811804524159694721%3A1h7btspnxtu > > > > If you can't find an answer, you could also boot DHCP and just use the > > kickstart to configure the static configuration, or you could > > also use DHCP reservations for IP assignment (which cobbler makes quite > > easy to handle with manage_dhcp enabled). > > > > > > > Jens Ahrens wrote: > > > > > >> I'm running FC7 with > > >> kernel-xen-2.6.20-2931.fc7 > > >> xen-libs-3.1.0-2.fc7 > > >> xen-3.1.0-2.fc7 > > >> cobbler-0.6.1-2.fc7 > > >> > > >> I'm trying to install a paravirtual machine > > >> (Scientific Linux 4.5) with a fixed ip-address(no > > >> private ip) and hostname and a kickstartfile > > >> accessible via e.g. http (nfs would also be possible). > > >> Everything seems to work, but the installation tries > > >> to determine the IP for eth0, which ends up in a menu > > >> asking for ip, netmask, gateway and dns. > > >> If I do the input manually the installtion continue > > >> using the values provided by my kickstart file. > > >> > > >> How can I pass the values like ip, netmask, gateway > > >> and dns to the SL4.5 installer, without having to type > > >> them in manually? > > >> > > >> Thanks in advance > > >> > > >> > > >> > > >> Wissenswertes f?r Bastler und Hobby Handwerker. BE A BETTER > > HEIMWERKER! www.yahoo.de/clever > > >> > > >> _______________________________________________ > > >> 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 > > > > -------------- next part -------------- An HTML attachment was scrubbed... URL: From dlutter at redhat.com Mon Sep 17 18:15:15 2007 From: dlutter at redhat.com (David Lutterkort) Date: Mon, 17 Sep 2007 11:15:15 -0700 Subject: [et-mgmt-tools] Can't make the meeting tomorrow In-Reply-To: <1190051761.20285.2.camel@localhost.localdomain> References: <1190051761.20285.2.camel@localhost.localdomain> Message-ID: <1190052915.20285.8.camel@localhost.localdomain> On Mon, 2007-09-17 at 17:56 +0000, David Lutterkort wrote: > I will be at IDF, having lots of fun. That, obviously, went to the wrong list. David From cam at cs.ualberta.ca Mon Sep 17 18:20:40 2007 From: cam at cs.ualberta.ca (Cameron Macdonell) Date: Mon, 17 Sep 2007 12:20:40 -0600 Subject: [et-mgmt-tools] virtmanager error when trying to connect on Ubuntu Message-ID: <1A230C56-581D-42E1-9944-D66971C21832@cs.ualberta.ca> Hi, I'm running KVM on Ubuntu and trying to get virtmanager running to manage them. I installed virt-manager and other than a gnome warning, it brings up the interface. I can see a single entry which is my host machine name. When I right-click and select "Connect", I get a python error saying: Traceback (most recent call last): File "/usr/local/share/virt-manager/virtManager/engine.py", line 205, in _do_show_connect self.show_connect() File "/usr/local/share/virt-manager/virtManager/engine.py", line 251, in show_connect self.windowConnect = vmmConnect(self.get_config(), self) File "/usr/local/share/virt-manager/virtManager/connect.py", line 54, in __init__ default = virtinst.util.default_connection() AttributeError: 'module' object has no attribute 'default_connection' Does anyone know what the problem is? Am I missing a package? Thanks, Cam From hbrock at redhat.com Mon Sep 17 18:43:59 2007 From: hbrock at redhat.com (Hugh Brock) Date: Mon, 17 Sep 2007 14:43:59 -0400 Subject: [et-mgmt-tools] virtmanager error when trying to connect on Ubuntu In-Reply-To: <1A230C56-581D-42E1-9944-D66971C21832@cs.ualberta.ca> References: <1A230C56-581D-42E1-9944-D66971C21832@cs.ualberta.ca> Message-ID: <46EECAEF.1060201@redhat.com> Cameron Macdonell wrote: > > Hi, > > I'm running KVM on Ubuntu and trying to get virtmanager running to > manage them. I installed virt-manager and other than a gnome warning, > it brings up the interface. I can see a single entry which is my host > machine name. When I right-click and select "Connect", I get a python > error saying: > > > Traceback (most recent call last): > File "/usr/local/share/virt-manager/virtManager/engine.py", line 205, > in _do_show_connect > self.show_connect() > File "/usr/local/share/virt-manager/virtManager/engine.py", line 251, > in show_connect > self.windowConnect = vmmConnect(self.get_config(), self) > File "/usr/local/share/virt-manager/virtManager/connect.py", line 54, > in __init__ > default = virtinst.util.default_connection() > AttributeError: 'module' object has no attribute 'default_connection' > > > Does anyone know what the problem is? Am I missing a package? > > Thanks, > Cam You'll need the very latest python-virtinst package as well -- it fixes this bug. python-virtinst is linked from virt-manager.org, you'll need to get the tip of the current tree in Mercurial. Good luck, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From cam at cs.ualberta.ca Mon Sep 17 19:26:45 2007 From: cam at cs.ualberta.ca (Cameron Macdonell) Date: Mon, 17 Sep 2007 13:26:45 -0600 Subject: [et-mgmt-tools] virtmanager error when trying to connect on Ubuntu In-Reply-To: <46EECAEF.1060201@redhat.com> References: <1A230C56-581D-42E1-9944-D66971C21832@cs.ualberta.ca> <46EECAEF.1060201@redhat.com> Message-ID: On 17-Sep-07, at 12:43 PM, Hugh Brock wrote: > Cameron Macdonell wrote: >> Hi, >> I'm running KVM on Ubuntu and trying to get virtmanager running to >> manage them. I installed virt-manager and other than a gnome >> warning, it brings up the interface. I can see a single entry >> which is my host machine name. When I right-click and select >> "Connect", I get a python error saying: >> Traceback (most recent call last): >> File "/usr/local/share/virt-manager/virtManager/engine.py", line >> 205, in _do_show_connect >> self.show_connect() >> File "/usr/local/share/virt-manager/virtManager/engine.py", line >> 251, in show_connect >> self.windowConnect = vmmConnect(self.get_config(), self) >> File "/usr/local/share/virt-manager/virtManager/connect.py", >> line 54, in __init__ >> default = virtinst.util.default_connection() >> AttributeError: 'module' object has no attribute 'default_connection' >> Does anyone know what the problem is? Am I missing a package? >> Thanks, >> Cam > > You'll need the very latest python-virtinst package as well -- it > fixes this bug. python-virtinst is linked from virt-manager.org, > you'll need to get the tip of the current tree in Mercurial. > > Good luck, > --Hugh > Ok, so that got rid of the error, but I still can't see any VMs. Is there some method by which I need to specify the VMs to virt-manager or should it detect any VMs that are running automatically? Thanks, Cam ----------------------------------------------- A. Cameron Macdonell Ph.D. Student Department of Computing Science University of Alberta cam at cs.ualberta.ca From hbrock at redhat.com Mon Sep 17 20:01:25 2007 From: hbrock at redhat.com (Hugh Brock) Date: Mon, 17 Sep 2007 16:01:25 -0400 Subject: [et-mgmt-tools] virtmanager error when trying to connect on Ubuntu In-Reply-To: References: <1A230C56-581D-42E1-9944-D66971C21832@cs.ualberta.ca> <46EECAEF.1060201@redhat.com> Message-ID: <46EEDD15.6060206@redhat.com> Cameron Macdonell wrote: > On 17-Sep-07, at 12:43 PM, Hugh Brock wrote: > >> Cameron Macdonell wrote: >>> Hi, >>> I'm running KVM on Ubuntu and trying to get virtmanager running to >>> manage them. I installed virt-manager and other than a gnome >>> warning, it brings up the interface. I can see a single entry which >>> is my host machine name. When I right-click and select "Connect", I >>> get a python error saying: >>> Traceback (most recent call last): >>> File "/usr/local/share/virt-manager/virtManager/engine.py", line >>> 205, in _do_show_connect >>> self.show_connect() >>> File "/usr/local/share/virt-manager/virtManager/engine.py", line >>> 251, in show_connect >>> self.windowConnect = vmmConnect(self.get_config(), self) >>> File "/usr/local/share/virt-manager/virtManager/connect.py", line >>> 54, in __init__ >>> default = virtinst.util.default_connection() >>> AttributeError: 'module' object has no attribute 'default_connection' >>> Does anyone know what the problem is? Am I missing a package? >>> Thanks, >>> Cam >> >> You'll need the very latest python-virtinst package as well -- it >> fixes this bug. python-virtinst is linked from virt-manager.org, >> you'll need to get the tip of the current tree in Mercurial. >> >> Good luck, >> --Hugh >> > > Ok, so that got rid of the error, but I still can't see any VMs. Is > there some method by which I need to specify the VMs to virt-manager or > should it detect any VMs that are running automatically? > > Thanks, > Cam > Well, it depends on how you have defined any VMs you are running. If you have defined KVM vms via libvirt, then virt-manager will get the relevant information for them from libvirt and display them in the UI. If you haven't defined any domains at all, you can create them by clicking the "new" button next to the KVM connection in virt-manager. If you've created a KVM guest outside of libvirt, you'll need to redefine it using libvirt via "virsh define", or manually place a complete XML domain definition in the correct directory under /etc/libvirt -- otherwise libvirt has no way of knowing anything about the guest and can't manage it for you. Hope this helps, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From tdsivakumar at gmail.com Tue Sep 18 05:30:56 2007 From: tdsivakumar at gmail.com (siva kumar kumar) Date: Tue, 18 Sep 2007 05:30:56 +0000 (GMT) Subject: [et-mgmt-tools] Link to call me for free Message-ID: <16750470.1262301190093456389.JavaMail.tomcat@media1.jaxtr.com> I am using jaxtr, and if you also sign up, we can talk for free on the phone at any time. -siva kumar P.S. Here is the link to sign up: http://www.jaxtr.com/user/ticket?n=T1rx7d8967g596&type=joininvite --- Delivered by jaxtr, Inc., 855 Oak Grove Avenue, Suite 100, Menlo Park, California 94025. To stop receiving messages from this sender go to http://www.jaxtr.com/user/reportabuse.jsp?it=T1rx7d8967g596 -------------- next part -------------- An HTML attachment was scrubbed... URL: From cam at cs.ualberta.ca Tue Sep 18 16:54:28 2007 From: cam at cs.ualberta.ca (Cam Macdonell) Date: Tue, 18 Sep 2007 10:54:28 -0600 Subject: [et-mgmt-tools] virtmanager error when trying to connect on Ubuntu In-Reply-To: <46EEDD15.6060206@redhat.com> References: <1A230C56-581D-42E1-9944-D66971C21832@cs.ualberta.ca> <46EECAEF.1060201@redhat.com> <46EEDD15.6060206@redhat.com> Message-ID: <46F002C4.40002@cs.ualberta.ca> Hugh Brock wrote: > Cameron Macdonell wrote: >> On 17-Sep-07, at 12:43 PM, Hugh Brock wrote: > Well, it depends on how you have defined any VMs you are running. If you > have defined KVM vms via libvirt, then virt-manager will get the > relevant information for them from libvirt and display them in the UI. > If you haven't defined any domains at all, you can create them by > clicking the "new" button next to the KVM connection in virt-manager. > > If you've created a KVM guest outside of libvirt, you'll need to > redefine it using libvirt via "virsh define", or manually place a > complete XML domain definition in the correct directory under > /etc/libvirt -- otherwise libvirt has no way of knowing anything about > the guest and can't manage it for you. > > Hope this helps, > --Hugh > Thanks Hugh, your help has been invaluable. I can start the VM, but I can't specify a VNC option (SDL works, though). I keep getting "Could not parse VNC address". I've read some pages that point to a dependency on gtkvnc error, but I wouldn't expect a dependency to manifest itself as an error like this (but I could be wrong). I'm using Ubuntu which does not package virt-manager, so I've had to build mostly from scratch, is there a config option I'm missing? Thanks again, Cam From hbrock at redhat.com Tue Sep 18 17:32:52 2007 From: hbrock at redhat.com (Hugh Brock) Date: Tue, 18 Sep 2007 13:32:52 -0400 Subject: [et-mgmt-tools] virtmanager error when trying to connect on Ubuntu In-Reply-To: <46F002C4.40002@cs.ualberta.ca> References: <1A230C56-581D-42E1-9944-D66971C21832@cs.ualberta.ca> <46EECAEF.1060201@redhat.com> <46EEDD15.6060206@redhat.com> <46F002C4.40002@cs.ualberta.ca> Message-ID: <46F00BC4.9020100@redhat.com> Cam Macdonell wrote: > Hugh Brock wrote: >> Cameron Macdonell wrote: >>> On 17-Sep-07, at 12:43 PM, Hugh Brock wrote: >> Well, it depends on how you have defined any VMs you are running. If >> you have defined KVM vms via libvirt, then virt-manager will get the >> relevant information for them from libvirt and display them in the UI. >> If you haven't defined any domains at all, you can create them by >> clicking the "new" button next to the KVM connection in virt-manager. >> >> If you've created a KVM guest outside of libvirt, you'll need to >> redefine it using libvirt via "virsh define", or manually place a >> complete XML domain definition in the correct directory under >> /etc/libvirt -- otherwise libvirt has no way of knowing anything about >> the guest and can't manage it for you. >> >> Hope this helps, >> --Hugh >> > > Thanks Hugh, your help has been invaluable. > > I can start the VM, but I can't specify a VNC option (SDL works, > though). I keep getting "Could not parse VNC address". I've read some > pages that point to a dependency on gtkvnc error, but I wouldn't expect > a dependency to manifest itself as an error like this (but I could be > wrong). I'm using Ubuntu which does not package virt-manager, so I've > had to build mostly from scratch, is there a config option I'm missing? > > Thanks again, > Cam Ahh yeah sorry this is a lot more difficult without packaging. I believe you are now running afoul of our newly introduced dependency on gtk-vnc, which is a gtk widget that renders a vnc connection. You can download it from http://gtk-vnc.sourceforge.net. Where is the error showing up exactly? Even without the gtk-vnc widget, you should be able to make a normal vnc connection to your guest using vncviewer (for example). --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From cam at cs.ualberta.ca Tue Sep 18 19:51:04 2007 From: cam at cs.ualberta.ca (Cam Macdonell) Date: Tue, 18 Sep 2007 13:51:04 -0600 Subject: [et-mgmt-tools] virtmanager error when trying to connect on Ubuntu In-Reply-To: <46F00BC4.9020100@redhat.com> References: <1A230C56-581D-42E1-9944-D66971C21832@cs.ualberta.ca> <46EECAEF.1060201@redhat.com> <46EEDD15.6060206@redhat.com> <46F002C4.40002@cs.ualberta.ca> <46F00BC4.9020100@redhat.com> Message-ID: <46F02C28.5090902@cs.ualberta.ca> Hugh Brock wrote: > Cam Macdonell wrote: >> Hugh Brock wrote: >>> Cameron Macdonell wrote: >>>> On 17-Sep-07, at 12:43 PM, Hugh Brock wrote: >>> Well, it depends on how you have defined any VMs you are running. If >>> you have defined KVM vms via libvirt, then virt-manager will get the >>> relevant information for them from libvirt and display them in the >>> UI. If you haven't defined any domains at all, you can create them by >>> clicking the "new" button next to the KVM connection in virt-manager. >>> >>> If you've created a KVM guest outside of libvirt, you'll need to >>> redefine it using libvirt via "virsh define", or manually place a >>> complete XML domain definition in the correct directory under >>> /etc/libvirt -- otherwise libvirt has no way of knowing anything >>> about the guest and can't manage it for you. >>> >>> Hope this helps, >>> --Hugh >>> >> >> Thanks Hugh, your help has been invaluable. >> >> I can start the VM, but I can't specify a VNC option (SDL works, >> though). I keep getting "Could not parse VNC address". I've read >> some pages that point to a dependency on gtkvnc error, but I wouldn't >> expect a dependency to manifest itself as an error like this (but I >> could be wrong). I'm using Ubuntu which does not package >> virt-manager, so I've had to build mostly from scratch, is there a >> config option I'm missing? >> >> Thanks again, >> Cam > > Ahh yeah sorry this is a lot more difficult without packaging. Yes, Fedora seems to be popular for most KVM users :-) > I believe > you are now running afoul of our newly introduced dependency on gtk-vnc, > which is a gtk widget that renders a vnc connection. You can download > it from http://gtk-vnc.sourceforge.net. I'll give it a shot. > Where is the error showing up exactly? Even without the gtk-vnc widget, > you should be able to make a normal vnc connection to your guest using > vncviewer (for example). > > --Hugh > The VM won't start at all. Nothing seems to point to gtk-vnc. The libvirtd output is: libvir: QEMU error : QEMU quit during console startup Could not parse VNC address Virt-manager pops up a console with this error: virDomainCreate() failed QEMU quit during console startup Could not parse VNC addresssudo echo 1 >> /proc/sys/net/ipv4/ip_forward sudo /sbin/iptables -t nat -A POSTROUTING -s 192.168.21.150 -j MASQUERADE the "Details" window within the popup provides this: Unable to start virtual machine ' virDomainCreate() failed QEMU quit during console startup Could not parse VNC address Traceback (most recent call last): File "/usr/local/share/virt-manager/virtManager/console.py", line 402, in control_vm_run self.vm.startup() File "/usr/local/share/virt-manager/virtManager/domain.py", line 377, in startup self.vm.create() File "/usr/local/lib/python2.5/site-packages/libvirt.py", line 228, in create if ret == -1: raise libvirtError ('virDomainCreate() failed', dom=self) libvirtError: virDomainCreate() failed QEMU quit during console startup Could not parse VNC address My graphics line in my xml is: I've also tried values 1 and -1 for port. Thanks again, Cam From hbrock at redhat.com Tue Sep 18 20:00:02 2007 From: hbrock at redhat.com (Hugh Brock) Date: Tue, 18 Sep 2007 16:00:02 -0400 Subject: [et-mgmt-tools] virtmanager error when trying to connect on Ubuntu In-Reply-To: <46F02C28.5090902@cs.ualberta.ca> References: <1A230C56-581D-42E1-9944-D66971C21832@cs.ualberta.ca> <46EECAEF.1060201@redhat.com> <46EEDD15.6060206@redhat.com> <46F002C4.40002@cs.ualberta.ca> <46F00BC4.9020100@redhat.com> <46F02C28.5090902@cs.ualberta.ca> Message-ID: <46F02E42.5040807@redhat.com> Cam Macdonell wrote: > > > The VM won't start at all. Nothing seems to point to gtk-vnc. > > The libvirtd output is: > libvir: QEMU error : QEMU quit during console startup > Could not parse VNC address > > Virt-manager pops up a console with this error: > > virDomainCreate() failed QEMU quit during console startup > Could not parse VNC addresssudo echo 1 >> /proc/sys/net/ipv4/ip_forward > sudo /sbin/iptables -t nat -A POSTROUTING -s 192.168.21.150 -j MASQUERADE > > the "Details" window within the popup provides this: > > Unable to start virtual machine ' > virDomainCreate() failed QEMU quit during console startup > Could not parse VNC address > > Traceback (most recent call last): > File "/usr/local/share/virt-manager/virtManager/console.py", line 402, > in control_vm_run > self.vm.startup() > File "/usr/local/share/virt-manager/virtManager/domain.py", line 377, > in startup > self.vm.create() > File "/usr/local/lib/python2.5/site-packages/libvirt.py", line 228, in > create > if ret == -1: raise libvirtError ('virDomainCreate() failed', dom=self) > libvirtError: virDomainCreate() failed QEMU quit during console startup > Could not parse VNC address > > My graphics line in my xml is: > > > > I've also tried values 1 and -1 for port. > That's pretty strange, particularly > virDomainCreate() failed QEMU quit during console startup > Could not parse VNC addresssudo echo 1 >> /proc/sys/net/ipv4/ip_forward > sudo /sbin/iptables -t nat -A POSTROUTING -s 192.168.21.150 -j MASQUERADE Can you post your entire domain XML and anything relevant at the end of /root/.virt-manager/virt-manager.log please? Could you also try starting the guest from the command line with "virsh start " and post the output from that? Thanks, --Hugh > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From berrange at redhat.com Tue Sep 18 20:02:43 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Tue, 18 Sep 2007 21:02:43 +0100 Subject: [et-mgmt-tools] virtmanager error when trying to connect on Ubuntu In-Reply-To: <46F02C28.5090902@cs.ualberta.ca> References: <1A230C56-581D-42E1-9944-D66971C21832@cs.ualberta.ca> <46EECAEF.1060201@redhat.com> <46EEDD15.6060206@redhat.com> <46F002C4.40002@cs.ualberta.ca> <46F00BC4.9020100@redhat.com> <46F02C28.5090902@cs.ualberta.ca> Message-ID: <20070918200243.GL19431@redhat.com> On Tue, Sep 18, 2007 at 01:51:04PM -0600, Cam Macdonell wrote: > > The VM won't start at all. Nothing seems to point to gtk-vnc. > > The libvirtd output is: > libvir: QEMU error : QEMU quit during console startup > Could not parse VNC address This most likely means your regular 'qemu' binary is older than the 'kvm' qemu binary. libvirt has an assumption that 'qemu' and 'kvm' are both based off the same base QEMU version. Since KVM is based off 0.9.0, you need to make sure your regular QEMU is also 0.9.0. If it is older, say, 0.8.x then this is the kind of error you'd get. This is because the VNC syntax for QEMU changed. 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 cam at cs.ualberta.ca Tue Sep 18 21:17:00 2007 From: cam at cs.ualberta.ca (Cam Macdonell) Date: Tue, 18 Sep 2007 15:17:00 -0600 Subject: [et-mgmt-tools] virtmanager error when trying to connect on Ubuntu In-Reply-To: <46F02E42.5040807@redhat.com> References: <1A230C56-581D-42E1-9944-D66971C21832@cs.ualberta.ca> <46EECAEF.1060201@redhat.com> <46EEDD15.6060206@redhat.com> <46F002C4.40002@cs.ualberta.ca> <46F00BC4.9020100@redhat.com> <46F02C28.5090902@cs.ualberta.ca> <46F02E42.5040807@redhat.com> Message-ID: <46F0404C.7070406@cs.ualberta.ca> Hugh Brock wrote: > That's pretty strange, particularly > > > virDomainCreate() failed QEMU quit during console startup > > Could not parse VNC addresssudo echo 1 >> /proc/sys/net/ipv4/ip_forward > > sudo /sbin/iptables -t nat -A POSTROUTING -s 192.168.21.150 -j > MASQUERADE > Ahh, my mistake, reckless cutting and pasting. The error was this: virDomainCreate() failed QEMU quit during console startup Could not parse VNC address > Can you post your entire domain XML and anything relevant at the end of > /root/.virt-manager/virt-manager.log please? Here's my XML domain. I grabbed it off of libvirt's documentation. I'm not sure what uuid should be. If you could explain that, that would be great too. cam at madhatter:~/KVM/feisty$ cat kvm-feisty.xml kvm-feisty 4dea24b3-1d52-d8f3-2516-782e98a23fa0 393216 1 hvm /usr/local/kvm/bin/qemu-system-x86_64 > > Could you also try starting the guest from the command line with "virsh > start " and post the output from that? It's the same error from the command line: cam at madhatter:~/KVM/feisty$ sudo virsh -c qemu:///system start kvm-feisty Password: libvir: QEMU error : QEMU quit during console startup Could not parse VNC address error: Failed to start domain kvm-feisty libvirtd spits out the same error as before: virDomainCreate() failed QEMU quit during console startup Could not parse VNC address Thanks again, Cam From cam at cs.ualberta.ca Tue Sep 18 21:24:27 2007 From: cam at cs.ualberta.ca (Cam Macdonell) Date: Tue, 18 Sep 2007 15:24:27 -0600 Subject: [et-mgmt-tools] virtmanager error when trying to connect on Ubuntu In-Reply-To: <20070918200243.GL19431@redhat.com> References: <1A230C56-581D-42E1-9944-D66971C21832@cs.ualberta.ca> <46EECAEF.1060201@redhat.com> <46EEDD15.6060206@redhat.com> <46F002C4.40002@cs.ualberta.ca> <46F00BC4.9020100@redhat.com> <46F02C28.5090902@cs.ualberta.ca> <20070918200243.GL19431@redhat.com> Message-ID: <46F0420B.6090004@cs.ualberta.ca> Daniel P. Berrange wrote: > On Tue, Sep 18, 2007 at 01:51:04PM -0600, Cam Macdonell wrote: >> The VM won't start at all. Nothing seems to point to gtk-vnc. >> >> The libvirtd output is: >> libvir: QEMU error : QEMU quit during console startup >> Could not parse VNC address > > This most likely means your regular 'qemu' binary is older than the > 'kvm' qemu binary. libvirt has an assumption that 'qemu' and 'kvm' > are both based off the same base QEMU version. Since KVM is based off > 0.9.0, you need to make sure your regular QEMU is also 0.9.0. If it > is older, say, 0.8.x then this is the kind of error you'd get. This > is because the VNC syntax for QEMU changed. > Hi, I build KVM from scratch, I'm on the latest build kvm-40. I'm not sure when the VNC syntax changed. Actually, that would make for a useful error message if it printed what syntax it's trying to use! One thing is that KVM doesn't use a port for VNC but outputs to a "display". Thanks, Cam From bkearney at redhat.com Wed Sep 19 12:28:11 2007 From: bkearney at redhat.com (Bryan Kearney) Date: Wed, 19 Sep 2007 08:28:11 -0400 Subject: [et-mgmt-tools] Alternative storage backends for Cobbler / other features In-Reply-To: <46D88E85.3050904@redhat.com> References: <46D88E85.3050904@redhat.com> Message-ID: <46F115DB.6000209@redhat.com> Couple of things having been using the cobbler cli for a while: 1) I find myself adding single ksmeta tags, and having to re-create the whole list. I would be nice to be able to CRUD a single element in the ksmeta field. 2) Much of the data in the profiles is "important" and I would like to put that into source code control. So.. a way to import and export a tree of profiles would be swanky. -- bk Michael DeHaan wrote: > So there's been a lot of interest around having Cobbler read LDAP > recently, and possibly I'm guessing it would be useful to have it > /write/ to LDAP. I know several > folks already have their own scripts to interface between the two. > > I'm thinking about abstracting the serializer code to allow for > configs in arbitrary formats, though the existing YAML will be the > default > and will not require any configuration for existing installs -- or new > ones. I will not taking your non-XML human-readable config files > away from you :) > Anyhow, the serializer stuff is already somewhat modular so I don't > expect this to be terribly complicated. The hard part will be > engineering things to not > need to worry about schema upgrades. > > The DB options are mainly to keep queries fast as we scale up into > thousands of system records. Future scaling work may also (probably) > imply looking more towards OMAPI when dealing with ISC's dhcp.conf > versus having to template out the file. If someone things that is > needed (or even better, would like to work on that), please speak up. > > Most likely what would happen is I'll implement the framework for > allowing arbitrary formats with a sqlite prototype, and if someone > else wants to add > in LDAP later that would be pretty easy to do by following the sqlite > module's lead. > > The other thing on the radar is finally making the XMLRPC API > bi-directional (by adding an additional secure version on another > port) to make the life of webapps using the Cobbler API easier. I've > been meaning to do that for a while. Until then apps that need write > access to cobbler configs can go through the python API and/or the > YAML tree. > > There was also a great suggestion about giving koan a very basic > --register function, that would add an entry in the cobbler DB that > tells the admin that he needs to set the system up. It would fill in > the MAC, IP, and the hostname -- but that's it. This is probably > going to be a bit further down the pipe than the above but it would be > useful for > cases where taking manual inventory of all the MACs in a datacenter > would be a bit painful. We already have most of the code to do this > from virt-factory, in fact, and we > can add this into the existing Cobbler API. > > Comments? Questions? Ideas? > > --Michael > > > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From kunkichi at gmail.com Wed Sep 19 17:01:25 2007 From: kunkichi at gmail.com (Shimizu Kuniaki) Date: Thu, 20 Sep 2007 02:01:25 +0900 Subject: [et-mgmt-tools] cobbler.et.redhat.com is down? Message-ID: <4dd11c330709191001oede0a31g44487789cf339bc4@mail.gmail.com> Hi, for the last few days, I've benn able to access cobbler.et.redhat.com. any problems on cobbler.et.redhat.com? Are there any mirrors from where I can download the latest source archive? From berrange at redhat.com Wed Sep 19 20:56:03 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 19 Sep 2007 21:56:03 +0100 Subject: [et-mgmt-tools] cobbler.et.redhat.com is down? In-Reply-To: <4dd11c330709191001oede0a31g44487789cf339bc4@mail.gmail.com> References: <4dd11c330709191001oede0a31g44487789cf339bc4@mail.gmail.com> Message-ID: <20070919205603.GM2712@redhat.com> On Thu, Sep 20, 2007 at 02:01:25AM +0900, Shimizu Kuniaki wrote: > for the last few days, I've benn able to access cobbler.et.redhat.com. > any problems on cobbler.et.redhat.com? Sorry for the inconvenience, we had some issues with the server, but it is back online now. 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 Thu Sep 20 11:41:35 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Thu, 20 Sep 2007 12:41:35 +0100 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.7 released Message-ID: <46F25C6F.8030903@redhat.com> We're pleased to announce the release of virt-top 0.3.2.7. virt-top is a top-like utility for showing stats of virtualized domains. Many keys and command line options are the same as for ordinary top. It uses libvirt so it capable of showing stats across a variety of different virtualization systems. Release 0.3.2.7 includes support for network and block device screens (press '2' and '3' keys respectively). We have also fixed a number of bugs. Source and binaries are available here: http://et.redhat.com/~rjones/virt-top/download.html Screenshots from here: http://et.redhat.com/~rjones/virt-top/screenshots.html If you're using Fedora 7/8, then enable the development repository and you should be able to just 'yum install virt-top'. 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 lfarkas at bppiac.hu Thu Sep 20 12:30:02 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Thu, 20 Sep 2007 14:30:02 +0200 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.7 released In-Reply-To: <46F25C6F.8030903@redhat.com> References: <46F25C6F.8030903@redhat.com> Message-ID: <46F267CA.3010501@bppiac.hu> Richard W.M. Jones wrote: > We're pleased to announce the release of virt-top 0.3.2.7. > > virt-top is a top-like utility for showing stats of virtualized domains. > Many keys and command line options are the same as for ordinary top. > > It uses libvirt so it capable of showing stats across a variety of > different virtualization systems. > > Release 0.3.2.7 includes support for network and block device screens > (press '2' and '3' keys respectively). We have also fixed a number of > bugs. > > Source and binaries are available here: > http://et.redhat.com/~rjones/virt-top/download.html the download page says: "Note that i386 binaries also work on x86-64", but when i try to install: ---------------------------------------- rpm -Uvh virt-top-0.3.2.7-1.i386.rpm error: Failed dependencies: libc.so.6 is needed by virt-top-0.3.2.7-1.i386 libc.so.6(GLIBC_2.0) is needed by virt-top-0.3.2.7-1.i386 libc.so.6(GLIBC_2.1) is needed by virt-top-0.3.2.7-1.i386 libc.so.6(GLIBC_2.1.2) is needed by virt-top-0.3.2.7-1.i386 libc.so.6(GLIBC_2.2) is needed by virt-top-0.3.2.7-1.i386 libc.so.6(GLIBC_2.3) is needed by virt-top-0.3.2.7-1.i386 libdl.so.2 is needed by virt-top-0.3.2.7-1.i386 libdl.so.2(GLIBC_2.0) is needed by virt-top-0.3.2.7-1.i386 libdl.so.2(GLIBC_2.1) is needed by virt-top-0.3.2.7-1.i386 libm.so.6 is needed by virt-top-0.3.2.7-1.i386 libm.so.6(GLIBC_2.0) is needed by virt-top-0.3.2.7-1.i386 libncurses.so.5 is needed by virt-top-0.3.2.7-1.i386 libvirt.so.0 is needed by virt-top-0.3.2.7-1.i386 ---------------------------------------- and of course i'd not like to isntall i[36]86 rpms. so it'd be useful to provides x86_64 rpms too for rhel5. thanks. -- Levente "Si vis pacem para bellum!" From rjones at redhat.com Thu Sep 20 12:36:27 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Thu, 20 Sep 2007 13:36:27 +0100 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.7 released In-Reply-To: <46F267CA.3010501@bppiac.hu> References: <46F25C6F.8030903@redhat.com> <46F267CA.3010501@bppiac.hu> Message-ID: <46F2694B.1020802@redhat.com> Farkas Levente wrote: > Richard W.M. Jones wrote: >> We're pleased to announce the release of virt-top 0.3.2.7. >> >> virt-top is a top-like utility for showing stats of virtualized domains. >> Many keys and command line options are the same as for ordinary top. >> >> It uses libvirt so it capable of showing stats across a variety of >> different virtualization systems. >> >> Release 0.3.2.7 includes support for network and block device screens >> (press '2' and '3' keys respectively). We have also fixed a number of >> bugs. >> >> Source and binaries are available here: >> http://et.redhat.com/~rjones/virt-top/download.html > > the download page says: "Note that i386 binaries also work on x86-64", > but when i try to install: > ---------------------------------------- > rpm -Uvh virt-top-0.3.2.7-1.i386.rpm > error: Failed dependencies: > libc.so.6 is needed by virt-top-0.3.2.7-1.i386 > libc.so.6(GLIBC_2.0) is needed by virt-top-0.3.2.7-1.i386 > libc.so.6(GLIBC_2.1) is needed by virt-top-0.3.2.7-1.i386 > libc.so.6(GLIBC_2.1.2) is needed by virt-top-0.3.2.7-1.i386 > libc.so.6(GLIBC_2.2) is needed by virt-top-0.3.2.7-1.i386 > libc.so.6(GLIBC_2.3) is needed by virt-top-0.3.2.7-1.i386 > libdl.so.2 is needed by virt-top-0.3.2.7-1.i386 > libdl.so.2(GLIBC_2.0) is needed by virt-top-0.3.2.7-1.i386 > libdl.so.2(GLIBC_2.1) is needed by virt-top-0.3.2.7-1.i386 > libm.so.6 is needed by virt-top-0.3.2.7-1.i386 > libm.so.6(GLIBC_2.0) is needed by virt-top-0.3.2.7-1.i386 > libncurses.so.5 is needed by virt-top-0.3.2.7-1.i386 > libvirt.so.0 is needed by virt-top-0.3.2.7-1.i386 Yes, you need the i386 libraries (glibc, ncurses and libvirt) since the i386 binary is linked to i386 libraries. Nevertheless, I'll make some 64 bit native RHEL 5 binaries in a moment. > and of course i'd not like to isntall i[36]86 rpms. so it'd be useful to > provides x86_64 rpms too for rhel5. 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 lfarkas at bppiac.hu Thu Sep 20 12:49:03 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Thu, 20 Sep 2007 14:49:03 +0200 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.7 released In-Reply-To: <46F2694B.1020802@redhat.com> References: <46F25C6F.8030903@redhat.com> <46F267CA.3010501@bppiac.hu> <46F2694B.1020802@redhat.com> Message-ID: <46F26C3F.9010005@bppiac.hu> Richard W.M. Jones wrote: > Farkas Levente wrote: >> Richard W.M. Jones wrote: >>> We're pleased to announce the release of virt-top 0.3.2.7. >>> >>> virt-top is a top-like utility for showing stats of virtualized domains. >>> Many keys and command line options are the same as for ordinary top. >>> >>> It uses libvirt so it capable of showing stats across a variety of >>> different virtualization systems. >>> >>> Release 0.3.2.7 includes support for network and block device screens >>> (press '2' and '3' keys respectively). We have also fixed a number of >>> bugs. >>> >>> Source and binaries are available here: >>> http://et.redhat.com/~rjones/virt-top/download.html >> >> the download page says: "Note that i386 binaries also work on x86-64", >> but when i try to install: >> ---------------------------------------- >> rpm -Uvh virt-top-0.3.2.7-1.i386.rpm >> error: Failed dependencies: >> libc.so.6 is needed by virt-top-0.3.2.7-1.i386 >> libc.so.6(GLIBC_2.0) is needed by virt-top-0.3.2.7-1.i386 >> libc.so.6(GLIBC_2.1) is needed by virt-top-0.3.2.7-1.i386 >> libc.so.6(GLIBC_2.1.2) is needed by virt-top-0.3.2.7-1.i386 >> libc.so.6(GLIBC_2.2) is needed by virt-top-0.3.2.7-1.i386 >> libc.so.6(GLIBC_2.3) is needed by virt-top-0.3.2.7-1.i386 >> libdl.so.2 is needed by virt-top-0.3.2.7-1.i386 >> libdl.so.2(GLIBC_2.0) is needed by virt-top-0.3.2.7-1.i386 >> libdl.so.2(GLIBC_2.1) is needed by virt-top-0.3.2.7-1.i386 >> libm.so.6 is needed by virt-top-0.3.2.7-1.i386 >> libm.so.6(GLIBC_2.0) is needed by virt-top-0.3.2.7-1.i386 >> libncurses.so.5 is needed by virt-top-0.3.2.7-1.i386 >> libvirt.so.0 is needed by virt-top-0.3.2.7-1.i386 > > Yes, you need the i386 libraries (glibc, ncurses and libvirt) since the > i386 binary is linked to i386 libraries. > > Nevertheless, I'll make some 64 bit native RHEL 5 binaries in a moment. and at the same time the src.rpm would be useful too (fedora/development dir is always a few hours/days behind you). -- Levente "Si vis pacem para bellum!" From rjones at redhat.com Thu Sep 20 13:13:38 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Thu, 20 Sep 2007 14:13:38 +0100 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.7 released In-Reply-To: <46F26C3F.9010005@bppiac.hu> References: <46F25C6F.8030903@redhat.com> <46F267CA.3010501@bppiac.hu> <46F2694B.1020802@redhat.com> <46F26C3F.9010005@bppiac.hu> Message-ID: <46F27202.50005@redhat.com> I've put up x86-64 (64 bit) binaries compiled on RHEL 5. http://et.redhat.com/~rjones/virt-top/download.html Note that you _will_ need libvirt >= 0.2.1, and that can be a problem if you are using RHEL 5 GA because I believe that shipped with libvirt 0.1.8. We require features from libvirt >= 0.2.1 so it's not likely that we'll ever backport to an older version. Source RPMs are here. On RHEL 5 GA they need to be compiled in the order: dnsmasq, libvirt, ocaml, ocaml-findlib, ocaml-extlib, ocaml-curses, ocaml-csv, ocaml-xml-light, ocaml-libvirt (this will become much simpler in future): http://www.annexia.org/tmp/r5/ 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 rjones at redhat.com Thu Sep 20 13:20:58 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Thu, 20 Sep 2007 14:20:58 +0100 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.7 released In-Reply-To: <46F27202.50005@redhat.com> References: <46F25C6F.8030903@redhat.com> <46F267CA.3010501@bppiac.hu> <46F2694B.1020802@redhat.com> <46F26C3F.9010005@bppiac.hu> <46F27202.50005@redhat.com> Message-ID: <46F273BA.9060202@redhat.com> Richard W.M. Jones wrote: > Note that you _will_ need libvirt >= 0.2.1, and that can be a problem if > you are using RHEL 5 GA because I believe that shipped with libvirt > 0.1.8. We require features from libvirt >= 0.2.1 so it's not likely > that we'll ever backport to an older version. I should add that you can just install libvirt >= 0.2.1 by any method (compile from source, RPM, etc.) and then install the virt-top binary. The binary is dynamically linked in such a way that it will work with any libvirt.so from 0.2.1 through to the very latest libvirt from CVS. 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 lfarkas at bppiac.hu Thu Sep 20 13:23:48 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Thu, 20 Sep 2007 15:23:48 +0200 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.7 released In-Reply-To: <46F27202.50005@redhat.com> References: <46F25C6F.8030903@redhat.com> <46F267CA.3010501@bppiac.hu> <46F2694B.1020802@redhat.com> <46F26C3F.9010005@bppiac.hu> <46F27202.50005@redhat.com> Message-ID: <46F27464.7070909@bppiac.hu> Richard W.M. Jones wrote: > I've put up x86-64 (64 bit) binaries compiled on RHEL 5. > > http://et.redhat.com/~rjones/virt-top/download.html > > Note that you _will_ need libvirt >= 0.2.1, and that can be a problem if > you are using RHEL 5 GA because I believe that shipped with libvirt > 0.1.8. We require features from libvirt >= 0.2.1 so it's not likely > that we'll ever backport to an older version. > > Source RPMs are here. On RHEL 5 GA they need to be compiled in the > order: dnsmasq, libvirt, ocaml, ocaml-findlib, ocaml-extlib, > ocaml-curses, ocaml-csv, ocaml-xml-light, ocaml-libvirt (this will > become much simpler in future): > > http://www.annexia.org/tmp/r5/ i already recompile all dependencies from fedora for the previous version, but while the previso version works on rhel5 this doesn't seems: ------------------------------------- [root at kvm ~]# /usr/bin/virt-top --connect qemu:///system libvir: error : this function is not supported by the hypervisor: virConnectGetHostname Fatal error: exception Libvirt.Virterror(3, 0, _, 2, _, 0, _, _, 0, _, _, 0) [root at kvm ~]# rpm -q libvirt libvirt-0.3.2-2 ------------------------------------- ps. and anyway the src.rpm for virt-top-0.3.2.7-1 still missing:-( -- Levente "Si vis pacem para bellum!" From rjones at redhat.com Thu Sep 20 13:28:22 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Thu, 20 Sep 2007 14:28:22 +0100 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.7 released In-Reply-To: <46F27464.7070909@bppiac.hu> References: <46F25C6F.8030903@redhat.com> <46F267CA.3010501@bppiac.hu> <46F2694B.1020802@redhat.com> <46F26C3F.9010005@bppiac.hu> <46F27202.50005@redhat.com> <46F27464.7070909@bppiac.hu> Message-ID: <46F27576.7050002@redhat.com> Farkas Levente wrote: > Richard W.M. Jones wrote: >> I've put up x86-64 (64 bit) binaries compiled on RHEL 5. >> >> http://et.redhat.com/~rjones/virt-top/download.html >> >> Note that you _will_ need libvirt >= 0.2.1, and that can be a problem if >> you are using RHEL 5 GA because I believe that shipped with libvirt >> 0.1.8. We require features from libvirt >= 0.2.1 so it's not likely >> that we'll ever backport to an older version. >> >> Source RPMs are here. On RHEL 5 GA they need to be compiled in the >> order: dnsmasq, libvirt, ocaml, ocaml-findlib, ocaml-extlib, >> ocaml-curses, ocaml-csv, ocaml-xml-light, ocaml-libvirt (this will >> become much simpler in future): >> >> http://www.annexia.org/tmp/r5/ > > i already recompile all dependencies from fedora for the previous > version, but while the previso version works on rhel5 this doesn't seems: > ------------------------------------- > [root at kvm ~]# /usr/bin/virt-top --connect qemu:///system > libvir: error : this function is not supported by the hypervisor: > virConnectGetHostname > Fatal error: exception Libvirt.Virterror(3, 0, _, 2, _, 0, _, _, 0, _, _, 0) > [root at kvm ~]# rpm -q libvirt > libvirt-0.3.2-2 You've found a real bug. Virt-top is not very well tested against qemu, or rather that should be not tested at all. But I'll go and try to make it work with qemu now. Probably take a few hours though ... > ps. and anyway the src.rpm for virt-top-0.3.2.7-1 still missing:-( The source is inside ocaml-libvirt srpm. 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 rjones at redhat.com Thu Sep 20 14:09:11 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Thu, 20 Sep 2007 15:09:11 +0100 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.7 released In-Reply-To: <46F27576.7050002@redhat.com> References: <46F25C6F.8030903@redhat.com> <46F267CA.3010501@bppiac.hu> <46F2694B.1020802@redhat.com> <46F26C3F.9010005@bppiac.hu> <46F27202.50005@redhat.com> <46F27464.7070909@bppiac.hu> <46F27576.7050002@redhat.com> Message-ID: <46F27F07.8010705@redhat.com> Richard W.M. Jones wrote: > Farkas Levente wrote: >> [root at kvm ~]# /usr/bin/virt-top --connect qemu:///system >> libvir: error : this function is not supported by the hypervisor: >> virConnectGetHostname >> Fatal error: exception Libvirt.Virterror(3, 0, _, 2, _, 0, _, _, 0, _, >> _, 0) >> [root at kvm ~]# rpm -q libvirt >> libvirt-0.3.2-2 > > You've found a real bug. Virt-top is not very well tested against qemu, > or rather that should be not tested at all. But I'll go and try to make > it work with qemu now. Probably take a few hours though ... The patch to fix that particular bug is attached. There may be other though, still testing ... 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: virt-top-qemu-hostname.patch Type: text/x-patch Size: 519 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 3237 bytes Desc: S/MIME Cryptographic Signature URL: From mdehaan at redhat.com Thu Sep 20 14:53:47 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 20 Sep 2007 10:53:47 -0400 Subject: [et-mgmt-tools] Alternative storage backends for Cobbler / other features In-Reply-To: <46F115DB.6000209@redhat.com> References: <46D88E85.3050904@redhat.com> <46F115DB.6000209@redhat.com> Message-ID: <46F2897B.7020407@redhat.com> Bryan Kearney wrote: > Couple of things having been using the cobbler cli for a while: > > 1) I find myself adding single ksmeta tags, and having to re-create > the whole list. I would be nice to be able to CRUD a single element in > the ksmeta field. Yep, I'm aware of this. The new Web UI will allow you to just edit the contents of the field (I'm due to send an email update out on that). The problem with adding or removing just a single tag from the command line was mainly in me not thinking of a great syntax to do this. It's easy enough to do adds but deletes get clunky. If you are careful, you can edit the yaml in /var/lib/cobbler/* directly ... though yaml is not super-forgiving of formatting errors. > 2) Much of the data in the profiles is "important" and I would like to > put that into source code control. So.. a way to import and export a > tree of profiles would be swanky. > Thanks for the bringing this up. I've thought about modifying the serializers to store seperate objects in seperate files, a la "conf.d". That would help more with the diffs. Otherwise, you can stick /var/lib/cobbler/(settings|profiles|distros|systems) in source control as is now. Though I do agree having support for a conf.d style structure would make this a lot better. Something to look at for 0.6.3, perhaps? > -- bk From mdehaan at redhat.com Thu Sep 20 15:09:10 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 20 Sep 2007 11:09:10 -0400 Subject: [et-mgmt-tools] New Cobbler Web UI about done ... testers/comments? Message-ID: <46F28D16.3000801@redhat.com> Al Tobey and I have been working on a Web UI for Cobbler. https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebUi We're still polishing up the last few remaining details, though it would be a great time for other folks to take a look, make comments, report bugs, and so on. This will help get release 0.6.2 out faster. From a git checkout, just run "make" and install the noarch RPM out of the "rpm-build" directory. The instructions on setting up the WebUI are at the link above. You'll see you are able to add cobbler objects (Distros, Profiles, Repos, and Systems) through the WebUI, edit existing ones, as well as run a "cobbler sync". There's the equivalent of a "cobbler list" in there also -- to see all that you have configured. Each field has comments below it saying what the field is for, frequently with examples. Where it a field has limited valid values, we're using radio boxes, selects, and so forth. So, we hope you find it easy to use. Suggestions are welcome, as I can see that we'll probably be adding to it some over time. A few known issues: -- If your login times out (30 minutes of no activity), and you can't log back in, try deleting your cookies from the boot server. Should be resolved soon. -- Kickstart template editing through the Web UI isn't there yet. We're working on that. Also, if you were interested in Cobbler's read-write XMLRPC API, the Web UI source code is an excellent example of this. 0.6.2 also includes Apache proxying all XMLRPC to port 80 (no more firewall configuration for koan), plus the ability to select multiple virtual disk locations for virt machines. There's also support for using Cobbler and DHCP management with multiple subnets, which I described in a previous email. A copy of that on the Cobbler Wiki is located here: https://hosted.fedoraproject.org/projects/cobbler/wiki/UsingCobblerDhcpManagementWithMultipleSubnets . So, there's other stuff to check out in 0.6.2 as well. If you have any Cobbler tips of your own, the Wiki is a great place to add them. You'll need to set up a Fedora account at https://admin.fedoraproject.org/accounts if you don't already have one and you'll be good to go. --Michael From pnixon at gmail.com Thu Sep 20 17:07:13 2007 From: pnixon at gmail.com (Patrick Nixon) Date: Thu, 20 Sep 2007 13:07:13 -0400 Subject: [et-mgmt-tools] cobbler webui issue Message-ID: Hey, Just git'd the new version with the webui, built the rpm without issue, updated the settings and auth.conf, and I'm getting an error when I try to connect. Did I miss a step somewhere? 2007-09-20 17:04:52,828 - cobbler.webui - INFO - login failed for: dcops 2007-09-20 17:04:52,829 - cobbler.webui - INFO - Exception occured: xmlrpclib.ProtocolError 2007-09-20 17:04:52,829 - cobbler.webui - INFO - Exception value: 2007-09-20 17:04:52,830 - cobbler.webui - INFO - Exception Info: File "/usr/lib/python2.4/site-packages/cobbler/webui/CobblerWeb.py", line 91, in __xmlrpc_setup self.token = self.remote.login( self.username, self.password ) File "/usr/lib/python2.4/xmlrpclib.py", line 1096, in __call__ return self.__send(self.__name, args) File "/usr/lib/python2.4/xmlrpclib.py", line 1383, in __request verbose=self.__verbose File "/usr/lib/python2.4/xmlrpclib.py", line 1137, in request headers I don't know python so I can't help diagnose the error, sorry :| --Pat From mdehaan at redhat.com Thu Sep 20 22:52:38 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 20 Sep 2007 18:52:38 -0400 Subject: [et-mgmt-tools] cobbler webui issue In-Reply-To: References: Message-ID: <46F2F9B6.6010106@redhat.com> Patrick Nixon wrote: > Hey, > Just git'd the new version with the webui, built the rpm without > issue, updated the settings and auth.conf, and I'm getting an error > when I try to connect. Did I miss a step somewhere? > The mailservers are slow today :) We diagnosed this one on IRC and basically all that was needed was to restart Apache once installed (due to the new cobbler.conf). --Michael > 2007-09-20 17:04:52,828 - cobbler.webui - INFO - login failed for: dcops > 2007-09-20 17:04:52,829 - cobbler.webui - INFO - Exception occured: > xmlrpclib.ProtocolError > 2007-09-20 17:04:52,829 - cobbler.webui - INFO - Exception value: > > 2007-09-20 17:04:52,830 - cobbler.webui - INFO - Exception Info: > File "/usr/lib/python2.4/site-packages/cobbler/webui/CobblerWeb.py", > line 91, in __xmlrpc_setup > self.token = self.remote.login( self.username, self.password ) > File "/usr/lib/python2.4/xmlrpclib.py", line 1096, in __call__ > return self.__send(self.__name, args) > File "/usr/lib/python2.4/xmlrpclib.py", line 1383, in __request > verbose=self.__verbose > File "/usr/lib/python2.4/xmlrpclib.py", line 1137, in request > headers > > > I don't know python so I can't help diagnose the error, sorry :| > > --Pat > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From pnixon at gmail.com Thu Sep 20 18:37:19 2007 From: pnixon at gmail.com (Patrick Nixon) Date: Thu, 20 Sep 2007 14:37:19 -0400 Subject: [et-mgmt-tools] XenEnterprise, PXEBoot, and Koan.... Message-ID: Hey all, I'm trying to figure out the best way to be able to pxeboot a Virtual Machine on our XenEnterprise server and have it build itself using cobbler. Since Xen doesn't do PXEBoot, I was reading up on Koan and the livecd thinking I could make an ISO for Xen to boot from and then koan to the cobbler server and 'Presto!' it's done. However, I'm having trouble building the LiveCD, but I think that's mostly because I'm using Centos5 instead for FC7+. Anyways, is a good viable solution or am I going about this completely wrong? Thanks for the feedback, --Pat From rjones at redhat.com Fri Sep 21 09:05:11 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Fri, 21 Sep 2007 10:05:11 +0100 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.7 released In-Reply-To: <46F27F07.8010705@redhat.com> References: <46F25C6F.8030903@redhat.com> <46F267CA.3010501@bppiac.hu> <46F2694B.1020802@redhat.com> <46F26C3F.9010005@bppiac.hu> <46F27202.50005@redhat.com> <46F27464.7070909@bppiac.hu> <46F27576.7050002@redhat.com> <46F27F07.8010705@redhat.com> Message-ID: <46F38947.9040709@redhat.com> Richard W.M. Jones wrote: > Richard W.M. Jones wrote: >> Farkas Levente wrote: >>> [root at kvm ~]# /usr/bin/virt-top --connect qemu:///system >>> libvir: error : this function is not supported by the hypervisor: >>> virConnectGetHostname >>> Fatal error: exception Libvirt.Virterror(3, 0, _, 2, _, 0, _, _, 0, >>> _, _, 0) >>> [root at kvm ~]# rpm -q libvirt >>> libvirt-0.3.2-2 >> >> You've found a real bug. Virt-top is not very well tested against >> qemu, or rather that should be not tested at all. But I'll go and try >> to make it work with qemu now. Probably take a few hours though ... > > The patch to fix that particular bug is attached. There may be other > though, still testing ... OK, I'm fairly sure this is the only patch needed. I'm going to roll this in to the next (0.3.2.8) release later. Libvirt needs works to support block and network stats from QEMU/KVM. Virt-top won't fail because this is missing, it just can't display anything in those columns. The work as far as I understand it is fairly simple, but also lower down on my list of priorities at the moment. Another thing which virt-top uses but which is missing in QEMU/KVM is physical CPU usage and CPU pinning. I have absolutely no idea if qemu supports this -- I assume it does through ordinary Linux mechanisms such as numactl. As for KVM, no idea whatsoever. 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 lfarkas at bppiac.hu Fri Sep 21 10:11:05 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Fri, 21 Sep 2007 12:11:05 +0200 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.7 released In-Reply-To: <46F38947.9040709@redhat.com> References: <46F25C6F.8030903@redhat.com> <46F267CA.3010501@bppiac.hu> <46F2694B.1020802@redhat.com> <46F26C3F.9010005@bppiac.hu> <46F27202.50005@redhat.com> <46F27464.7070909@bppiac.hu> <46F27576.7050002@redhat.com> <46F27F07.8010705@redhat.com> <46F38947.9040709@redhat.com> Message-ID: <46F398B9.5070707@bppiac.hu> Richard W.M. Jones wrote: > Richard W.M. Jones wrote: >> Richard W.M. Jones wrote: >>> Farkas Levente wrote: >>>> [root at kvm ~]# /usr/bin/virt-top --connect qemu:///system >>>> libvir: error : this function is not supported by the hypervisor: >>>> virConnectGetHostname >>>> Fatal error: exception Libvirt.Virterror(3, 0, _, 2, _, 0, _, _, 0, >>>> _, _, 0) >>>> [root at kvm ~]# rpm -q libvirt >>>> libvirt-0.3.2-2 >>> >>> You've found a real bug. Virt-top is not very well tested against >>> qemu, or rather that should be not tested at all. But I'll go and >>> try to make it work with qemu now. Probably take a few hours though ... >> >> The patch to fix that particular bug is attached. There may be other >> though, still testing ... > > OK, I'm fairly sure this is the only patch needed. > > I'm going to roll this in to the next (0.3.2.8) release later. > > Libvirt needs works to support block and network stats from QEMU/KVM. > Virt-top won't fail because this is missing, it just can't display > anything in those columns. The work as far as I understand it is fairly > simple, but also lower down on my list of priorities at the moment. > > Another thing which virt-top uses but which is missing in QEMU/KVM is > physical CPU usage and CPU pinning. I have absolutely no idea if qemu > supports this -- I assume it does through ordinary Linux mechanisms such > as numactl. As for KVM, no idea whatsoever. maybe you can ask kvm people:-) anyway if the new version will be uploaded i can try it again. -- Levente "Si vis pacem para bellum!" From pnixon at gmail.com Thu Sep 20 19:26:54 2007 From: pnixon at gmail.com (Patrick Nixon) Date: Thu, 20 Sep 2007 15:26:54 -0400 Subject: [et-mgmt-tools] Re: cobbler webui issue In-Reply-To: References: Message-ID: At Michael's advice from the IRC Channel, I've restarted the Apache server which cleared the issue up. --Pat On 9/20/07, Patrick Nixon wrote: > Hey, > Just git'd the new version with the webui, built the rpm without > issue, updated the settings and auth.conf, and I'm getting an error > when I try to connect. Did I miss a step somewhere? > > 2007-09-20 17:04:52,828 - cobbler.webui - INFO - login failed for: dcops > 2007-09-20 17:04:52,829 - cobbler.webui - INFO - Exception occured: > xmlrpclib.ProtocolError > 2007-09-20 17:04:52,829 - cobbler.webui - INFO - Exception value: > > 2007-09-20 17:04:52,830 - cobbler.webui - INFO - Exception Info: > File "/usr/lib/python2.4/site-packages/cobbler/webui/CobblerWeb.py", > line 91, in __xmlrpc_setup > self.token = self.remote.login( self.username, self.password ) > File "/usr/lib/python2.4/xmlrpclib.py", line 1096, in __call__ > return self.__send(self.__name, args) > File "/usr/lib/python2.4/xmlrpclib.py", line 1383, in __request > verbose=self.__verbose > File "/usr/lib/python2.4/xmlrpclib.py", line 1137, in request > headers > > > I don't know python so I can't help diagnose the error, sorry :| > > --Pat > From berrange at redhat.com Fri Sep 21 13:39:18 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Fri, 21 Sep 2007 14:39:18 +0100 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.7 released In-Reply-To: <46F38947.9040709@redhat.com> References: <46F25C6F.8030903@redhat.com> <46F267CA.3010501@bppiac.hu> <46F2694B.1020802@redhat.com> <46F26C3F.9010005@bppiac.hu> <46F27202.50005@redhat.com> <46F27464.7070909@bppiac.hu> <46F27576.7050002@redhat.com> <46F27F07.8010705@redhat.com> <46F38947.9040709@redhat.com> Message-ID: <20070921133918.GA17528@redhat.com> On Fri, Sep 21, 2007 at 10:05:11AM +0100, Richard W.M. Jones wrote: > Richard W.M. Jones wrote: > >Richard W.M. Jones wrote: > >>Farkas Levente wrote: > >>>[root at kvm ~]# /usr/bin/virt-top --connect qemu:///system > >>>libvir: error : this function is not supported by the hypervisor: > >>>virConnectGetHostname > >>>Fatal error: exception Libvirt.Virterror(3, 0, _, 2, _, 0, _, _, 0, > >>>_, _, 0) > >>>[root at kvm ~]# rpm -q libvirt > >>>libvirt-0.3.2-2 > >> > >>You've found a real bug. Virt-top is not very well tested against > >>qemu, or rather that should be not tested at all. But I'll go and try > >>to make it work with qemu now. Probably take a few hours though ... > > > >The patch to fix that particular bug is attached. There may be other > >though, still testing ... > > OK, I'm fairly sure this is the only patch needed. > > I'm going to roll this in to the next (0.3.2.8) release later. > > Libvirt needs works to support block and network stats from QEMU/KVM. > Virt-top won't fail because this is missing, it just can't display > anything in those columns. The work as far as I understand it is fairly > simple, but also lower down on my list of priorities at the moment. > > Another thing which virt-top uses but which is missing in QEMU/KVM is > physical CPU usage and CPU pinning. I have absolutely no idea if qemu > supports this -- I assume it does through ordinary Linux mechanisms such > as numactl. As for KVM, no idea whatsoever. QEMU is completely single process, even with SMP. So you can pin the QEMU process to a physical CPU & it'll pin every virtual CPU at once. KVM is different in that it uses one thread per vCPU, so you can pin individual threads to physical CPUs. 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 Sep 21 15:52:58 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 21 Sep 2007 11:52:58 -0400 Subject: [et-mgmt-tools] XenEnterprise, PXEBoot, and Koan.... In-Reply-To: References: Message-ID: <46F3E8DA.9080900@redhat.com> Patrick Nixon wrote: > Hey all, > I'm trying to figure out the best way to be able to pxeboot a Virtual > Machine on our XenEnterprise server and have it build itself using > cobbler. > > koan currently does Xen paravirt and also qemu/kqemu/KVM installs. So no Xen PXE as of yet. Upstream virtinst (the library koan uses) does have Xen PXE support, but as it's not released quite yet I haven't added it to koan. Of course, you can already use Cobbler to manage the PXE config, just that the piece to make koan create a new PXE virtual machine isn't in Cobbler yet. > Since Xen doesn't do PXEBoot, I was reading up on Koan and the livecd > thinking I could make an ISO for Xen to boot from and then koan to the > cobbler server and 'Presto!' it's done. > This really doesn't make sense -- and may indicate I haven't explained the point of the Live CD clearly enough That Live CD was designed for doing installs of hardware (bare metal) in an environment where Cobbler was present, but DHCP could not be configured. It's also a universal installer, meaning that F7 based live CD can install any distribution, not just those based on F-7. You can walk one CD anywhere on your network, stick it in, and have it install -- regardless of whether the admin of the network has PXE configured or not -- regardless of whether he's set up cobbler or not. It points that physical machine at /your/ cobbler server. It's just enough to boot it, discover the hard drives, and write grub and the neccessary details into the MBR. However, it's not really appropriate for doing virtualized installs. It assumes disks already exist and are discovered by the Boot CD, etc. Plus it wouldn't specify how much RAM you need, etc. Once PXE support is released in virtinst, I'll add some bits for koan to make it be able to install fullvirt Xen -- the PXE config will come from cobbler, as will the cobbler settings for how much RAM to allocate and where the disk space is. However, the actual boot will take place over PXE. Until then, you can use paravirtualized Xen or qemu/KVM with koan. They do not technically PXE but they achieve the same kind of result. > However, I'm having trouble building the LiveCD, but I think that's > mostly because I'm using Centos5 instead for FC7+. > The live CD must be built on F7 or higher. The only reason we require a live CD to be built (vs downloaded) is the parameters to koan (such as the boot server name, or the preference to autodiscover, and what profile to install) are embedded into the CD when it's burned. An interactive version that asks what profile to install from a list, however, is something that could be done with minimal changes. And if we did that, it could also prompt for the server location... From wright at imageworks.com Fri Sep 21 16:44:24 2007 From: wright at imageworks.com (Peter Wright) Date: Fri, 21 Sep 2007 09:44:24 -0700 Subject: [et-mgmt-tools] New Cobbler Web UI about done ... testers/comments? In-Reply-To: <46F28D16.3000801@redhat.com> References: <46F28D16.3000801@redhat.com> Message-ID: <46F3F4E8.5090401@imageworks.com> this is great! couple comments inline.... Michael DeHaan wrote: > > Al Tobey and I have been working on a Web UI for Cobbler. > > https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebUi > in the wiki page the URL to access the webui is reported as: https://bootserver.example.com/cgi-bin/cobbler/webui.cgi although on my system the correct URL seems to be: https://bootserver.example.com/cgi-bin/cobbler_webui.cgi > > You'll see you are able to add cobbler objects (Distros, Profiles, > Repos, and Systems) through the WebUI, edit existing ones, as well as > run a "cobbler sync". There's the equivalent of a "cobbler list" in > there also -- to see all that you have configured. Each field has > comments below it saying what the field is for, frequently with > examples. Where it a field has limited valid values, we're using radio > boxes, selects, and so forth. So, we hope you find it easy to use. > Suggestions are welcome, as I can see > that we'll probably be adding to it some over time. > > A few known issues: > -- If your login times out (30 minutes of no activity), and you > can't log back in, try deleting your cookies from the boot server. > Should be resolved soon. > -- Kickstart template editing through the Web UI isn't there yet. > We're working on that. > Everything looks excellent so far. So much so I feel that I can start handing this off to other admins in my group to assist with provisioning systems in our lab. after a quick checkout the only request i have would be if one could sort by Hostname in the system_list UI. We name our systems based on MAC address as well, so having the MAC field is redundant in our environment. Not sure about other folks though. once i had this off if i get any comments/suggests i'll be sure to let you all know. we'll be adding new systems and modifying existing systems so hopefully we can help out with the debugging. Other than that this looks awesome, thanks Al and Michael!! I'll owe both of you at least a beer if we ever meet up :) -pete -- Peter Wright Systems Administrator Sony Pictures Imageworks wright at imageworks.com www.imageworks.com From mdehaan at redhat.com Fri Sep 21 16:59:57 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 21 Sep 2007 12:59:57 -0400 Subject: [et-mgmt-tools] New Cobbler Web UI about done ... testers/comments? In-Reply-To: <46F3F4E8.5090401@imageworks.com> References: <46F28D16.3000801@redhat.com> <46F3F4E8.5090401@imageworks.com> Message-ID: <46F3F88D.9020402@redhat.com> Peter Wright wrote: > this is great! couple comments inline.... > > Michael DeHaan wrote: >> >> Al Tobey and I have been working on a Web UI for Cobbler. >> https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebUi >> > in the wiki page the URL to access the webui is reported as: > > https://bootserver.example.com/cgi-bin/cobbler/webui.cgi > > > > although on my system the correct URL seems to be: > > https://bootserver.example.com/cgi-bin/cobbler_webui.cgi It depends when you did the pull. I moved the directories around to namespace them yesterday I believe. > > > > > >> >> You'll see you are able to add cobbler objects (Distros, Profiles, >> Repos, and Systems) through the WebUI, edit existing ones, as well as >> run a "cobbler sync". There's the equivalent of a "cobbler list" in >> there also -- to see all that you have configured. Each field has >> comments below it saying what the field is for, frequently with >> examples. Where it a field has limited valid values, we're using radio >> boxes, selects, and so forth. So, we hope you find it easy to use. >> Suggestions are welcome, as I can see >> that we'll probably be adding to it some over time. >> >> A few known issues: >> -- If your login times out (30 minutes of no activity), and you >> can't log back in, try deleting your cookies from the boot server. >> Should be resolved soon. -- Kickstart template editing through >> the Web UI isn't there yet. We're working on that. >> > > > > Everything looks excellent so far. So much so I feel that I can start > handing this off to other admins in my group to assist with > provisioning systems in our lab. > after a quick checkout the only request i have would be if one could > sort by Hostname in the system_list UI. We name our systems based on > MAC address as well, so having the MAC field is redundant in our > environment. Not sure about other folks though. I think Al mentioned having some slick dynamic sorting code. I'll check up with him on that. > > once i had this off if i get any comments/suggests i'll be sure to let > you all know. we'll be adding new systems and modifying existing > systems so hopefully we can help out with the debugging. > > Other than that this looks awesome, thanks Al and Michael!! I'll owe > both of you at least a beer if we ever meet up :) Will have to take you up on that :) > > -pete > From hbrock at redhat.com Fri Sep 21 20:36:43 2007 From: hbrock at redhat.com (Hugh Brock) Date: Fri, 21 Sep 2007 16:36:43 -0400 Subject: [et-mgmt-tools] new cdrom connect/disconnect UI Message-ID: <46F42B5B.5010903@redhat.com> Hi everyone. I've just pushed a change that adds a connect/disconnect button next to the cdrom "source" line in the hardware details pane for full-virt guests. This means if your guest has a cdrom device, you can disconnect the "source" (backing) device for it, which allows you to change the cdrom or let another guest use it. When you need to use the device again, you can reconnect it using the same UI. This UI is a wrapper around libvirt's virDomainAttachDevice, which I recently modified to allow modifying a device in addition to attaching it. Enjoy, --Hugh -- Red Hat Virtualization Group http://redhat.com/virtualization Hugh Brock | virt-manager http://virt-manager.org hbrock at redhat.com | virtualization library http://libvirt.org From kunkichi at gmail.com Sat Sep 22 14:06:39 2007 From: kunkichi at gmail.com (Shimizu Kuniaki) Date: Sat, 22 Sep 2007 23:06:39 +0900 Subject: [et-mgmt-tools] "yum_repo_stanza" duplicates same "repo" entry on kickstart file. Message-ID: <4dd11c330709220706o1900c264k3138daad107b1d51@mail.gmail.com> Hi, Fist of all, this is great tools. good job, guys. Well, I have a question about "$yum_repo_stanza" and "$yum_config_stanza" on a kickstart template file. I went through the following steps: 1. Install cobbler. 2. Check and setup. 3. Import distro from CentOS5 DVD-ROM. # cobbler import --mirror=/mnt/cdrom --name="centos5" 2 distros and 2 profiles were automatically added. # cobbler report distro : centos5-i386 kernel : /var/www/cobbler/ks_mirror/centos5/images/pxeboot/vmlinuz initrd : /var/www/cobbler/ks_mirror/centos5/images/pxeboot/initrd.img kernel options : {} architecture : x86 ks metadata : {'tree': 'http://@@server@@/cblr/links/centos5-i386'} breed : redhat distro : centos5-xen-i386 kernel : /var/www/cobbler/ks_mirror/centos5/images/xen/vmlinuz initrd : /var/www/cobbler/ks_mirror/centos5/images/xen/initrd.img kernel options : {} architecture : x86 ks metadata : {'tree': 'http://@@server@@/cblr/links/centos5-xen-i386'} breed : redhat profile : centos5-i386 distro : centos5-i386 kickstart : /etc/cobbler/kickstart_fc6.ks kernel options : {} ks metadata : {} virt file size : 5 virt ram : 512 virt type : auto virt path : repos : [] profile : centos5-xen-i386 distro : centos5-xen-i386 kickstart : /etc/cobbler/kickstart_fc6.ks kernel options : {} ks metadata : {} virt file size : 5 virt ram : 512 virt type : auto virt path : repos : [] 4. Then, checked kickstart files in /var/www/cobbler/kickstarts/centos5-i386 # cat /var/www/cobbler/kickstarts/centos5-i386/ks.cfg --- repo --name=centos5 --baseurl=http://xxx.xxx.xxx.xxx/cobbler/ks_mirror/centos5 repo --name=centos5 --baseurl=http://xxx.xxx.xxx.xxx/cobbler/ks_mirror/centos5 repo --name=centos5 --baseurl=http://xxx.xxx.xxx.xxx/cobbler/ks_mirror/centos5 --- 5. Also, modified /var/lib/cobbler/settings, changed "yum_core_mirror_from_server" to "1", and sync. $yum_config_stanza was replaced like this. # cat /var/www/cobbler/kickstarts/centos5-i386/ks.cfg --- wget http://xxx.xxx.xxx.xxx/cobbler/ks_mirror/config/centos5-i386-0.repo --output-document=/etc/yum.repos.d/centos5-i386-0.repo wget http://xxx.xxx.xxx.xxx/cobbler/ks_mirror/config/centos5-i386-1.repo --output-document=/etc/yum.repos.d/centos5-i386-1.repo wget http://xxx.xxx.xxx.xxxcobbler/ks_mirror/config/centos5-i386-2.repo --output-document=/etc/yum.repos.d/centos5-i386-2.repo --- There are 3 repo setting files for i386 in /var/www/cobbler/ks_mirror/config # ls /var/www/cobbler/ks_mirror/config centos5-i386-0.repo centos5-i386-1.repo centos5-i386-2.repo And, each content of those file is almost same. # diff centos5-i386-0.repo centos5-i386-1.repo 1,2c1,2 < [core-0] < name=core-0 --- > [core-1] > name=core-1 I checked action_import.py and found something seemed to relate this. ----- # write a yum config file that shows how to use the repo. if counter == 0: dotrepo = "%s.repo" % distro.name else: dotrepo = "%s-%s.repo" % (distro.name, counter) ----- I think this is not harmless, but does mean nothing. But I don't know about python and I am not sure what this means. What does this means? Kuniaki Shimizu. ----- OS: CentOS5 Cobbler Version: 0.6.1-2 (build from srpm) From tobert at gmail.com Sun Sep 23 22:42:51 2007 From: tobert at gmail.com (Al Tobey) Date: Sun, 23 Sep 2007 15:42:51 -0700 Subject: [et-mgmt-tools] cobbler patch Message-ID: <5ac7acb10709231542q1fdbef2dr7577cf274d2f99fd@mail.gmail.com> Importing the FC7 DVD on FC7 fails for me if I use the automatically created/mounted mountpoint that has spaces in it (Fedora 7 x86_64 DVD). Simply adding quotes to action_import.py fixes the problem. -Al diff --git a/cobbler/action_import.py b/cobbler/action_import.py index b9aaf0e..267207a 100644 --- a/cobbler/action_import.py +++ b/cobbler/action_import.py @@ -26,7 +26,7 @@ import shutil from rhpl.translate import _, N_, textdomain, utf8 WGET_CMD = "wget --mirror --no-parent --no-host-directories --directory-prefix %s/%s %s" -RSYNC_CMD = "rsync -a %s %s %s/ks_mirror/%s --exclude-from=/etc/cobbler/rsync.exclude -- +RSYNC_CMD = "rsync -a %s '%s' %s/ks_mirror/%s --exclude-from=/etc/cobbler/rsync.exclude TRY_LIST = [ "Fedora", "Packages", "RedHat", "Client", "Server", "Centos", "CentOS", From mdehaan at redhat.com Mon Sep 24 14:40:06 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 24 Sep 2007 10:40:06 -0400 Subject: [et-mgmt-tools] "yum_repo_stanza" duplicates same "repo" entry on kickstart file. In-Reply-To: <4dd11c330709220706o1900c264k3138daad107b1d51@mail.gmail.com> References: <4dd11c330709220706o1900c264k3138daad107b1d51@mail.gmail.com> Message-ID: <46F7CC46.4060106@redhat.com> Shimizu Kuniaki wrote: > Hi, > > Fist of all, this is great tools. good job, guys. > > Well, I have a question about "$yum_repo_stanza" and > "$yum_config_stanza" on a kickstart template file. > > I went through the following steps: > > 1. Install cobbler. > > 2. Check and setup. > > 3. Import distro from CentOS5 DVD-ROM. > > ok... > 4. Then, checked kickstart files in /var/www/cobbler/kickstarts/centos5-i386 > > # cat /var/www/cobbler/kickstarts/centos5-i386/ks.cfg > --- > repo --name=centos5 > --baseurl=http://xxx.xxx.xxx.xxx/cobbler/ks_mirror/centos5 > repo --name=centos5 > --baseurl=http://xxx.xxx.xxx.xxx/cobbler/ks_mirror/centos5 > repo --name=centos5 > --baseurl=http://xxx.xxx.xxx.xxx/cobbler/ks_mirror/centos5 > --- > This doesn't look quite right, I'll investigate. I did a c5-i386 DVD import the other day (for testing) and only saw one repo show up (as intended). RHEL5, however, should show up as 3-4 repos, due to the ways the trees are split (seperate virtualization repo/tree on the DVD, etc). Looks like, at least in your case, Cobbler became slightly confused. > > I think this is not harmless, but does mean nothing. > But I don't know about python and I am not sure what this means. > It means the import most likely detected that you had a more complicated DVD than it really was :) It won't hurt anything, having the extra definitions, but yes, it would be better if they were not there. If you want to prune them manually just for now, look in /var/lib/cobbler/distros and remove the extra "source repos" from the list and that will fix the yum configuration. (I'd recommend backing up your /var/lib/cobbler files before doing this). > > > Kuniaki Shimizu. > > ----- > > OS: CentOS5 > Cobbler Version: 0.6.1-2 (build from srpm) > > _______________________________________________ > 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 Sep 24 14:43:06 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Mon, 24 Sep 2007 10:43:06 -0400 Subject: [et-mgmt-tools] cobbler patch In-Reply-To: <5ac7acb10709231542q1fdbef2dr7577cf274d2f99fd@mail.gmail.com> References: <5ac7acb10709231542q1fdbef2dr7577cf274d2f99fd@mail.gmail.com> Message-ID: <46F7CCFA.7000001@redhat.com> Al Tobey wrote: > Importing the FC7 DVD on FC7 fails for me if I use the automatically > created/mounted mountpoint that has spaces in it (Fedora 7 x86_64 > DVD). Simply adding quotes to action_import.py fixes the problem. > > -Al > > Applied, thanks! > diff --git a/cobbler/action_import.py b/cobbler/action_import.py > index b9aaf0e..267207a 100644 > --- a/cobbler/action_import.py > +++ b/cobbler/action_import.py > @@ -26,7 +26,7 @@ import shutil > from rhpl.translate import _, N_, textdomain, utf8 > > WGET_CMD = "wget --mirror --no-parent --no-host-directories > --directory-prefix %s/%s %s" > -RSYNC_CMD = "rsync -a %s %s %s/ks_mirror/%s > --exclude-from=/etc/cobbler/rsync.exclude -- > +RSYNC_CMD = "rsync -a %s '%s' %s/ks_mirror/%s > --exclude-from=/etc/cobbler/rsync.exclude > > TRY_LIST = [ > "Fedora", "Packages", "RedHat", "Client", "Server", "Centos", "CentOS", > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From rjones at redhat.com Mon Sep 24 16:10:47 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Mon, 24 Sep 2007 17:10:47 +0100 Subject: [et-mgmt-tools] ANNOUNCE: virt-top 0.3.2.8 released Message-ID: <46F7E187.2050008@redhat.com> We are pleased to announce the release of virt-top 0.3.2.8. virt-top is a top-like utility for showing stats of virtualized domains. Many keys and command line options are the same as for ordinary top. It uses libvirt so it capable of showing stats across a variety of different virtualization systems. Release 0.3.2.8 includes: - fixes so that you can monitor qemu and KVM domains - support for .virt-toprc init file, 'W' key to write an init file with the current settings - code tidy ups You can get source and binaries from: http://et.redhat.com/~rjones/virt-top/download.html Screenshots from here: http://et.redhat.com/~rjones/virt-top/screenshots.html If you're using Fedora 7/8, then enable the development repository and you should be able to just 'yum install virt-top'. 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 rjones at redhat.com Tue Sep 25 12:36:33 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Tue, 25 Sep 2007 13:36:33 +0100 Subject: [et-mgmt-tools] ANNOUNCE: nagios-virt: Nagios configuration tool for virtualisation Message-ID: <46F900D1.5020000@redhat.com> I'm pleased to announce the first public release of nagios-virt, which is a tool for automatically configuring Nagios to monitor virtual machines. http://et.redhat.com/~rjones/nagios-virt/ 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 kunkichi at gmail.com Tue Sep 25 14:27:14 2007 From: kunkichi at gmail.com (Shimizu Kuniaki) Date: Tue, 25 Sep 2007 23:27:14 +0900 Subject: [et-mgmt-tools] "yum_repo_stanza" duplicates same "repo" entry on kickstart file. In-Reply-To: <46F7CC46.4060106@redhat.com> References: <4dd11c330709220706o1900c264k3138daad107b1d51@mail.gmail.com> <46F7CC46.4060106@redhat.com> Message-ID: <4dd11c330709250727o79683089m6558eab912e75113@mail.gmail.com> Hi, Michael Thank you for reply. I fixed as you said, and yes! This is what I've expected. For People who encounter the same, I'll Show you how I fixed /var/lib/cobbler/distros and how things go. ----- --- - arch: x86 breed: redhat depth: 0 initrd: /var/www/cobbler/ks_mirror/centos5/images/xen/initrd.img kernel: /var/www/cobbler/ks_mirror/centos5/images/xen/vmlinuz kernel_options: {} ks_meta: tree: 'http://@@server@@/cblr/links/centos5-xen-i386' name: centos5-xen-i386 parent: ~ source_repos: - - 'http://@@server@@/cobbler/ks_mirror/config/centos5-xen-i386-0.repo' - 'http://@@server@@/cobbler/ks_mirror/centos5' - - 'http://@@server@@/cobbler/ks_mirror/config/centos5-xen-i386-1.repo' - 'http://@@server@@/cobbler/ks_mirror/centos5' - arch: x86 breed: redhat depth: 0 initrd: /var/www/cobbler/ks_mirror/centos5/images/pxeboot/initrd.img kernel: /var/www/cobbler/ks_mirror/centos5/images/pxeboot/vmlinuz kernel_options: {} ks_meta: tree: 'http://@@server@@/cblr/links/centos5-i386' name: centos5-i386 parent: ~ source_repos: - - 'http://@@server@@/cobbler/ks_mirror/config/centos5-i386-0.repo' - 'http://@@server@@/cobbler/ks_mirror/centos5' - - 'http://@@server@@/cobbler/ks_mirror/config/centos5-i386-1.repo' - 'http://@@server@@/cobbler/ks_mirror/centos5' - - 'http://@@server@@/cobbler/ks_mirror/config/centos5-i386-2.repo' - 'http://@@server@@/cobbler/ks_mirror/centos5' ----- Then, I fixed like this. ----- --- - arch: x86 breed: redhat depth: 0 initrd: /var/www/cobbler/ks_mirror/centos5/images/xen/initrd.img kernel: /var/www/cobbler/ks_mirror/centos5/images/xen/vmlinuz kernel_options: {} ks_meta: tree: 'http://@@server@@/cblr/links/centos5-xen-i386' name: centos5-xen-i386 parent: ~ source_repos: - - 'http://@@server@@/cobbler/ks_mirror/config/centos5-xen-i386-0.repo' - 'http://@@server@@/cobbler/ks_mirror/centos5' - arch: x86 breed: redhat depth: 0 initrd: /var/www/cobbler/ks_mirror/centos5/images/pxeboot/initrd.img kernel: /var/www/cobbler/ks_mirror/centos5/images/pxeboot/vmlinuz kernel_options: {} ks_meta: tree: 'http://@@server@@/cblr/links/centos5-i386' name: centos5-i386 parent: ~ source_repos: - - 'http://@@server@@/cobbler/ks_mirror/config/centos5-i386-0.repo' - 'http://@@server@@/cobbler/ks_mirror/centos5' ----- Cobber sync. The kickstart file was changed like this. ----- # If any cobbler repo definitions were referenced in the kickstart profile, include them here. *BLANK* *snip* %post wget http://XXX.XXX.XXX.XXX/cobbler/ks_mirror/config/centos5-i386-0.repo --output-document=/etc/yum.repos.d/centos5-i386-0.repo test -e /etc/yum.repos.d/fedora-core.repo && mv /etc/yum.repos.d/fedora-core.repo /etc/yum.repos.d/disabled-fedora-core test -e /etc/yum.repos.d/Centos-Base.repo && mv /etc/yum.repos.d/Centos-Base.repo /etc/yum.repos.d/disabled-Centos-Base ----- Good. Next, I added some UPDATES, EXTRAS, ADDONS repository. ----- #cobbler repo add --mirror=http://one-of-centos-mirrors/5.0/updates/i386/ --name=co5-i386-updates --local-filename=centos5-updates #cobbler repo add --mirror=http://one-of-centos-mirrors/centos/5.0/extras/i386/ --name=co5-i386-extras --local-filename=centos5-extras #cobbler repo add --mirror=http://one-of-centos-mirrors/centos/5.0/addons/i386/ --name=co5-i386-addons --local-filename=centos5-addons # cobbler reposync # cobbler sync # cobbler profile edit --name="centos5-i386" --repos="co5-i386-updates co5-i386-extras co5-i386-addons" # cobbler sync ----- Then, ks.cfg was changed like this. ----- # If any cobbler repo definitions were referenced in the kickstart profile, include them here. repo --name=co5-i386-updates --baseurl=http://XXX.XXX.XXX.XXX/cblr/repo_mirror/co5-i386-updates repo --name=co5-i386-extras --baseurl=http://XXX.XXX.XXX.XXX/cblr/repo_mirror/co5-i386-extras repo --name=co5-i386-addons --baseurl=http://XXX.XXX.XXX.XXX/cblr/repo_mirror/co5-i386-addons *snip* %post wget http://XXX.XXX.XXX.XXX/cblr/repo_mirror/co5-i386-updates/config.repo --output-document=/etc/yum.repos.d/centos5-updates.repo wget http://XXX.XXX.XXX.XXX/cblr/repo_mirror/co5-i386-extras/config.repo --output-document=/etc/yum.repos.d/centos5-extras.repo wget http://XXX.XXX.XXX.XXX/cblr/repo_mirror/co5-i386-addons/config.repo --output-document=/etc/yum.repos.d/centos5-addons.repo wget http://XXX.XXX.XXX.XXX/cobbler/ks_mirror/config/centos5-i386-0.repo --output-document=/etc/yum.repos.d/centos5-i386-0.repo test -e /etc/yum.repos.d/fedora-core.repo && mv /etc/yum.repos.d/fedora-core.repo /etc/yum.repos.d/disabled-fedora-core test -e /etc/yum.repos.d/Centos-Base.repo && mv /etc/yum.repos.d/Centos-Base.repo /etc/yum.repos.d/disabled-Centos-Base ----- Perfect! Thanks, Michael! From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Tue Sep 25 15:02:24 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Tue, 25 Sep 2007 17:02:24 +0200 Subject: [et-mgmt-tools] Two easy to fix bugs in cobbler Message-ID: <20070925170224.1df482cb@python3.es.egwn.lan> Hi, Both "bugs" are with cobbler-0.6.1-2 from Fedora. First one : Using "url --url $tree" in a kickstart file doesn't work with cobbler, as it wants the equal sign : "url --url=$tree"... but both are valid kickstart syntax, so cobbler should probably be fixed to allow both syntaxes. Second one : The /etc/logrotate.d/cobblerd_rotate file should be fixed : - Add "missingok" since otherwise a daily error is sent about no logs existing if cobbler was installed but has never been run. - Add ">/dev/null" to the restart line, to not get the useless output by email. Alternately, "condrestart" could be used instead of the test+restart, since it's implemented in the init script. If I should bugzilla these to keep track of them, please let me know. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.30 0.25 0.28 From lfarkas at bppiac.hu Tue Sep 25 15:07:27 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Tue, 25 Sep 2007 17:07:27 +0200 Subject: [et-mgmt-tools] ANNOUNCE: nagios-virt: Nagios configuration tool for virtualisation In-Reply-To: <46F900D1.5020000@redhat.com> References: <46F900D1.5020000@redhat.com> Message-ID: <46F9242F.4010706@bppiac.hu> Richard W.M. Jones wrote: > I'm pleased to announce the first public release of nagios-virt, which > is a tool for automatically configuring Nagios to monitor virtual machines. > > http://et.redhat.com/~rjones/nagios-virt/ an rpm always would be useful too;-) -- Levente "Si vis pacem para bellum!" From mdehaan at redhat.com Tue Sep 25 15:10:45 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 25 Sep 2007 11:10:45 -0400 Subject: [et-mgmt-tools] Two easy to fix bugs in cobbler In-Reply-To: <20070925170224.1df482cb@python3.es.egwn.lan> References: <20070925170224.1df482cb@python3.es.egwn.lan> Message-ID: <46F924F5.5060309@redhat.com> Matthias Saou wrote: > Hi, > > Both "bugs" are with cobbler-0.6.1-2 from Fedora. > > First one : > > Using "url --url $tree" in a kickstart file doesn't work with cobbler, > as it wants the equal sign : "url --url=$tree"... but both are valid > kickstart syntax, so cobbler should probably be fixed to allow both > syntaxes. > Cobbler accepts kickstart files as plaintext. What do you mean by "doesn't work with Cobbler"? It is true there's a regex in koan that needs to be tweaked to allow validation of the later, though I wasn't sure if you were talking about that or not. > Second one : > > The /etc/logrotate.d/cobblerd_rotate file should be fixed : > - Add "missingok" since otherwise a daily error is sent about no logs > existing if cobbler was installed but has never been run. > This is already added in git. > - Add ">/dev/null" to the restart line, to not get the useless output > by email. Alternately, "condrestart" could be used instead of the > test+restart, since it's implemented in the init script. > Will do. > If I should bugzilla these to keep track of them, please let me know. > Email is good enough. Thanks! > Matthias > > From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Tue Sep 25 15:19:53 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Tue, 25 Sep 2007 17:19:53 +0200 Subject: [et-mgmt-tools] Two easy to fix bugs in cobbler In-Reply-To: <46F924F5.5060309@redhat.com> References: <20070925170224.1df482cb@python3.es.egwn.lan> <46F924F5.5060309@redhat.com> Message-ID: <20070925171953.467e8c24@python3.es.egwn.lan> Michael DeHaan wrote : > > Using "url --url $tree" in a kickstart file doesn't work with cobbler, > > as it wants the equal sign : "url --url=$tree"... but both are valid > > kickstart syntax, so cobbler should probably be fixed to allow both > > syntaxes. > > > > Cobbler accepts kickstart files as plaintext. What do you mean by > "doesn't work with Cobbler"? > > It is true there's a regex in koan that needs to be tweaked to allow > validation of the later, though I wasn't > sure if you were talking about that or not. Oops, yes, I got confused, everything is good on cobbler's side, and the kickstart file validates, but koan reports that the url is missing from the kickstart file. It took me a while to figure out it was expecting the equal sign :-) Thanks for the rest! I'll probably try out a current snapshot since many of the issues I'm encountering seem to already be fixed ;-) Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.46 0.33 0.28 From karhong.ng at nsn.com Tue Sep 25 15:29:41 2007 From: karhong.ng at nsn.com (karhong.ng at nsn.com) Date: Tue, 25 Sep 2007 19:29:41 +0400 Subject: [et-mgmt-tools] Machine RHEL 5 reboots during creation of second guest OS Message-ID: > Hello, > > I have created sucessfully a fully virtualized guest OS (RHEL 5) > using virt-install. > I used a custom ISO image as the source to boot from. > It works. > > Then I try to creating a second fully virtualized OS (RHEL 5) as > above using the same method using virt-install. > During the creation of the second guest OS's HDD image file, it > caused the host (the physical machine) RHEL 5 to reboot. > > I also try using virt-manager to create the second fully > virtualized OS. The procedure also caused the physical machine to > reboot. > > I have attached the various files for your reference. > According to line# 108 of xend.log, it says the following : > ============== cut ============================ > [2007-09-25 14:58:41 xend 3410] DEBUG (DevController:110) > DevController: writing {'backend-id': '0', 'virtual-device': '768', > 'device-type': 'disk', 'state': '1', 'backend': > '/local/domain/0/backend/vbd/2/768'} to > /local/domain/2/device/vbd/768. > [2007-09-25 14:58:41 xend 3410] DEBUG (DevController:112) > DevController: writing {'domain': 'clas1', 'frontend': > '/local/domain/2/device/vbd/768', 'dev': 'hda', 'state': '1', > 'params': '/var/lib/xen/images/clas1_xen_image', 'mode': 'w', > 'online': '1', 'frontend-id': '2', 'type': 'file'} to > /local/domain/0/backend/vbd/2/768. > [2007-09-25 14:58:41 xend 3410] DEBUG (blkif:24) exception > looking up device number for hdc: [Errno 2] No such file or directory: > '/dev/hdc' > [2007-09-25 14:58:41 xend 3410] DEBUG (DevController:110) > DevController: writing {'backend-id': '0', 'virtual-device': '5632', > 'device-type': 'cdrom', 'state': '1', 'backend': > '/local/domain/0/backend/vbd/2/5632'} to > /local/domain/2/device/vbd/5632. > [2007-09-25 14:58:41 xend 3410] DEBUG (DevController:112) > DevController: writing {'domain': 'clas1', 'frontend': > '/local/domain/2/device/vbd/5632', 'dev': 'hdc', 'state': '1', > 'params': '/tmp/NPM31-OS-custom_kiskstart.iso', 'mode': 'r', 'online': > '1', 'frontend-id': '2', 'type': 'file'} to > /local/domain/0/backend/vbd/2/5632. > [2007-09-25 14:58:41 xend 3410] DEBUG (DevController:110) > DevController: writing {'state': '1', 'backend-id': '0', 'backend': > '/local/domain/0/backend/vif/2/0'} to /local/domain/2/device/vif/0. > ============= cut ============================= > > What could be wrong with my configurations ? > Kindly advice. Many thanks. > > Ps: > The below are the versions of software in the system : > Virtualization-en-US-5.0.0-7 > libvirt-0.1.8-15.el5 > virt-manager-0.2.6-7.el5 > python-virtinst-0.99.0-2.el5 > libvirt-python-0.1.8-15.el5 > > <> -------------- next part -------------- An HTML attachment was scrubbed... URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: xen_log.zip Type: application/x-zip-compressed Size: 33657 bytes Desc: xen_log.zip URL: From wright at imageworks.com Tue Sep 25 18:21:03 2007 From: wright at imageworks.com (Peter Wright) Date: Tue, 25 Sep 2007 11:21:03 -0700 Subject: [et-mgmt-tools] Two easy to fix bugs in cobbler In-Reply-To: <46F924F5.5060309@redhat.com> References: <46F924F5.5060309@redhat.com> Message-ID: <46F9518F.1000408@imageworks.com> Michael DeHaan wrote: > > Matthias Saou wrote: > > Hi, > > > > Both "bugs" are with cobbler-0.6.1-2 from Fedora. > > > > First one : > > > > Using "url --url $tree" in a kickstart file doesn't work with cobbler, > > as it wants the equal sign : "url --url=$tree"... but both are valid > > kickstart syntax, so cobbler should probably be fixed to allow both > > syntaxes. > > > > Cobbler accepts kickstart files as plaintext. What do you mean by > "doesn't work with Cobbler"? > > It is true there's a regex in koan that needs to be tweaked to allow > validation of the later, though I wasn't > sure if you were talking about that or not. > hmm...i believe this issue may prevent people from using NFS install media. it's not a big issue as i've just been lazy and have not updated my ks.cfg's to use http. -pete -- Peter Wright Systems Administrator Sony Pictures Imageworks wright at imageworks.com www.imageworks.com From mdehaan at redhat.com Tue Sep 25 19:21:05 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 25 Sep 2007 15:21:05 -0400 Subject: [et-mgmt-tools] Two easy to fix bugs in cobbler In-Reply-To: <46F9518F.1000408@imageworks.com> References: <46F924F5.5060309@redhat.com> <46F9518F.1000408@imageworks.com> Message-ID: <46F95FA1.6040200@redhat.com> >> > hmm...i believe this issue may prevent people from using NFS install > media. it's not a big issue as i've just been lazy and have not > updated my ks.cfg's to use http. > > -pete > > If you look into the evils that cobbler does when it sees NFS in a kickstart template, you'll see that it will insert a commented out URL line in for koan to find and understand. It's weird, and patches to make that cleaner are welcome, but it seems to work. So, as long as your kickstarts are stored on a filesystem such that they run through the templating engine, life should be good with koan. (This doesn't come into play with PXE at all) --Michael From mdehaan at redhat.com Tue Sep 25 22:48:10 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Tue, 25 Sep 2007 18:48:10 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release Message-ID: <46F9902A.8060205@redhat.com> It's new Cobbler & koan time. The new feature this time is the Cobbler Web UI, for which I have to give a big hand to Al Tobey for coding up a *massive* chunk of it. The WUI recently gained the ability to edit kickstart files and is really shiny (IMHO) at this point. While the rest of cobbler we'll consider "stable", the Web UI component should be considered a "beta" just for this 1st release. Definitely help test it out though, and report back your findings/ideas/comments. If you or fellow admins need to make tweaks from time to time and need a better way of visualizing your provisioning configuration, the WebUI is probably for you. As mentioned before, the setup instructions are on the Wiki, so just follow these steps once things are installed: https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebUi Unless you've been requesting a feature (in which case, it probably was added), the other changes are relatively minor and you might not notice them. I've sent emails out about most of them already. One thing admins should note is that /var/lib/cobbler/settings is finally packaged as a config file, so you'll get an "rpmnew" for the new settings. That will be moved into /etc probably in the next release and all will be right with config files :) So ... I'm uploading the 0.6.2 RPM's today for folks to have a chance to try things out before pushing to the mirrors on Friday. http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm FYI -- The changelogs might look rather long, though it's been about a month since the last release and lots of folks have been working off the source-control versions. So things are rather well tested... For Cobbler: - cobbler repo auto-add to discover yum repos automatically (requires a rather new yum) - fix bug that allows empty mac addresses (None) in dhcpd.conf - kickstarts automatically save kickstart file used to /root/cobbler.ks - allow multiple (comma-seperated) values for --virt-size - removed deprecated 'enchant' function (use SSH and koan instead) - cleanup of a few unused settings - allow for serialization modules to be selected in /etc/cobbler/modules.conf - patch to allow for reposync of specific repos, even if not set to update - added --dhcp-tag section for better DHCP customization (esp with multiple subnets) - added Apache proxying around XMLRPC port for wider network access - refactor XMLRPC API and establish a read-write API - allow for configuring of read-write XMLRPC users in /etc/cobbler/auth.conf - WebUI - packaged /var/lib/cobbler/settings as a config file - added BuildRequires to help build on other platforms - relocate cgi-bin files to cgi-bin/cobbler for namespacing - fix syslog logging for systems not in the cobbler DB. For koan - Allow for multiple values (comma seperated) to --virt-path - Allow for multiple virt file sizes - Refactoring around the virt bits - Remove legacy sys.path hack as enchant has been removed (just use SSH) - Try port 80 proxied XMLRPC by default (first) before trying 25151 As usual, a huge thanks to everyone who submitted patches, emails, or hangs out in #cobbler. --Michael From berrange at redhat.com Wed Sep 26 01:24:57 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Wed, 26 Sep 2007 02:24:57 +0100 Subject: [et-mgmt-tools] ANNOIUNCE: virt-manager/virt-install Wiki site Message-ID: <20070926012457.GL17213@redhat.com> The main http://virt-manager.org website is becoming home for a growing collection of management tools built around libvirt - virt-manager, virt-install, virt-clone and virt-image so far. The content on the site though is pretty static & not updated as often as I would like. So I decided to setup a Wiki instance under part of the site - notice the 'Wiki' link in the navigation area. I'll move things like the FAQ and development docs into the wiki so they're more easily updateble. Anonymous edits are not allowed, but user account registration is open to anyone who wishes to contribute to the Wiki. 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 thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Wed Sep 26 08:55:49 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Wed, 26 Sep 2007 10:55:49 +0200 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <46F9902A.8060205@redhat.com> References: <46F9902A.8060205@redhat.com> Message-ID: <20070926105549.3b9862c1@python3.es.egwn.lan> Michael DeHaan wrote : > So ... I'm uploading the 0.6.2 RPM's today for folks to have a chance to > try things out before pushing to the mirrors on Friday. > > http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm > http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm Great! I'll be testing them today hopefully. A few points regarding the packages : - Why does cobbler require "python-devel" at runtime? - Could you maybe put absolute URLs into the Source0 fields? - Cobbler seems to be missing an excludearch ppc64, which koan has. You might also want to put a comment in the spec file as to why they are needed, as I don't understand why cobbler couldn't be used on a ppc, myself. - The License field for koan needs to be updated too. - You should use "rm -rf $RPM_BUILD_ROOT". The variable will never be empty, so the pseudo-check is useless. - The .gz suffix to the man pages shouldn't be included, since rpm can be configured to use bzip2 or no compression, use foo.1*. - Inconsistency between "/sbin/chkconfig" and "chkconfig". - There are two Requires(preun) instead of having a Requires(postun). - Use "%ifarch %{ix86}" instead of "%ifarch i386 i686" to also include athlon, i586 etc. Probably more. If you don't mind an almost rewrite, I could also simply take a stab at making all the changes and submit updated spec files. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.39 0.30 0.27 From Kevin.Maute at WPAFB.AF.MIL Wed Sep 26 14:39:15 2007 From: Kevin.Maute at WPAFB.AF.MIL (Maute, Kevin P CTR AFRL/VAOO) Date: Wed, 26 Sep 2007 10:39:15 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <46F9902A.8060205@redhat.com> References: <46F9902A.8060205@redhat.com> Message-ID: <29A7C5F65D54744CBFD97CD62A33DEE9023848CB@VFOHMLAO13.Enterprise.afmc.ds.af.mil> Thanks Michael & Al! I am somewhat of a newbie also with xmlrpc and am having trouble authenticating to webui. I have selinux disabled and have "xmlrpc_rw_enabled: 1" and did a "service cobblerd restart" after updating to 0.6.2-1 from 0.6.1. I am unclear what "anyusername = yourpasswordhere" means. I tried the testuser with its password but did not have any joy. I also tried my existing user/password combo. I am assuming auth.conf is similar to .htaccess but am a little concerned about the clear text aspect. Thanks, Kevin > -----Original Message----- > From: et-mgmt-tools-bounces at redhat.com > [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of Michael DeHaan > Sent: Tuesday, September 25, 2007 6:48 PM > To: 'et-mgmt-tools at redhat.com' > Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release > > It's new Cobbler & koan time. > > The new feature this time is the Cobbler Web UI, for which I have to > give a big hand to Al Tobey for coding up a *massive* chunk of it. > The WUI recently gained the ability to edit kickstart files and is > really shiny (IMHO) at this point. While the rest of cobbler we'll > consider "stable", the Web UI component should be considered a "beta" > just for this 1st release. Definitely help test it out though, > and report back your findings/ideas/comments. If you or > fellow admins > need to make tweaks from time to time and need a better > way of visualizing your provisioning configuration, the WebUI is > probably for you. > > As mentioned before, the setup instructions are on the Wiki, so just > follow these steps once things are installed: > > https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebUi > > Unless you've been requesting a feature (in which case, it > probably was > added), the other changes are relatively minor > and you might not notice them. I've sent emails out about > most of them > already. One thing admins should note is that > /var/lib/cobbler/settings is finally packaged as a config file, so > you'll get an "rpmnew" for the new settings. That will > be moved into /etc probably in the next release and all will be right > with config files :) > > So ... I'm uploading the 0.6.2 RPM's today for folks to have > a chance to > try things out before pushing to the mirrors on Friday. > > http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm > http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm > > FYI -- The changelogs might look rather long, though it's > been about a > month since the last release and lots of folks have been > working off the source-control versions. So things are rather well > tested... > > For Cobbler: > - cobbler repo auto-add to discover yum repos automatically > (requires a > rather new yum) > - fix bug that allows empty mac addresses (None) in dhcpd.conf > - kickstarts automatically save kickstart file used to > /root/cobbler.ks > - allow multiple (comma-seperated) values for --virt-size > - removed deprecated 'enchant' function (use SSH and koan instead) > - cleanup of a few unused settings > - allow for serialization modules to be selected in > /etc/cobbler/modules.conf > - patch to allow for reposync of specific repos, even if not > set to update > - added --dhcp-tag section for better DHCP customization (esp with > multiple subnets) > - added Apache proxying around XMLRPC port for wider network access > - refactor XMLRPC API and establish a read-write API > - allow for configuring of read-write XMLRPC users in > /etc/cobbler/auth.conf > - WebUI > - packaged /var/lib/cobbler/settings as a config file > - added BuildRequires to help build on other platforms > - relocate cgi-bin files to cgi-bin/cobbler for namespacing > - fix syslog logging for systems not in the cobbler DB. > > For koan > - Allow for multiple values (comma seperated) to --virt-path > - Allow for multiple virt file sizes > - Refactoring around the virt bits > - Remove legacy sys.path hack as enchant has been removed > (just use SSH) > - Try port 80 proxied XMLRPC by default (first) before trying 25151 > > As usual, a huge thanks to everyone who submitted patches, emails, or > hangs out in #cobbler. > > --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 Wed Sep 26 14:57:34 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 26 Sep 2007 10:57:34 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <20070926105549.3b9862c1@python3.es.egwn.lan> References: <46F9902A.8060205@redhat.com> <20070926105549.3b9862c1@python3.es.egwn.lan> Message-ID: <46FA735E.7030400@redhat.com> Matthias Saou wrote: > Michael DeHaan wrote : > > >> So ... I'm uploading the 0.6.2 RPM's today for folks to have a chance to >> try things out before pushing to the mirrors on Friday. >> >> http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm >> http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm >> > > Great! I'll be testing them today hopefully. A few points regarding the > packages : > > - Why does cobbler require "python-devel" at runtime? > Woo, free RPM review :) BuildRequires is a requirement that is added in Fedora 7. The Requires can definitely be removed. > - Could you maybe put absolute URLs into the Source0 fields? > Sure. http://cobbler.et.redhat.com/download/* > - Cobbler seems to be missing an excludearch ppc64, which koan has. > Technically, it's possible to run a cobbler server on a ppc64 box. PXE may not work without copying a syslinux over, but koan would be fine. See comments below about %ifarch with noarch packages and syslinux. > - The License field for koan needs to be updated too. > You're right. Thanks. > - You should use "rm -rf $RPM_BUILD_ROOT". The variable will never > be empty, so the pseudo-check is useless. > That patch was committed by someone who apparently needed it, I'm not sure why. I'll pull it if there are no objections. I don't see a reason for keeping it. > - The .gz suffix to the man pages shouldn't be included, since rpm > can be configured to use bzip2 or no compression, use foo.1*. > Sounds fine, though FYI -- currently the Makefile explicitly builds them with gzip compression, therefore you'll never get the bzip2's. I'm not sure how far back man supported bzip2's, though space does not seem to be a huge concern. > - Inconsistency between "/sbin/chkconfig" and "chkconfig". > Easy to fix. > - There are two Requires(preun) instead of having a Requires(postun). > Good find. > - Use "%ifarch %{ix86}" instead of "%ifarch i386 i686" to also include > athlon, i586 etc. > %ifarch apparently doesn't work for NoArch packages, so it makes more sense to remove syslinux as a dependency and just have cobbler check notice if it's not there -- or to just exclude the arches we don't support. In fact, I thought I made this change already but apparently I didn't :) The alternative is to do the ExcludeArch on ppc. How many folks are running Cobbler servers on ppc? Probably almost none. > Probably more. If you don't mind an almost rewrite, I could also simply > take a stab at making all the changes and submit updated spec files. > Sure! If you can seperate them into a series of smaller patches, that would be great. git-send-email is awesome. > Matthias > > From mdehaan at redhat.com Wed Sep 26 15:04:16 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 26 Sep 2007 11:04:16 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <29A7C5F65D54744CBFD97CD62A33DEE9023848CB@VFOHMLAO13.Enterprise.afmc.ds.af.mil> References: <46F9902A.8060205@redhat.com> <29A7C5F65D54744CBFD97CD62A33DEE9023848CB@VFOHMLAO13.Enterprise.afmc.ds.af.mil> Message-ID: <46FA74F0.9050308@redhat.com> Maute, Kevin P CTR AFRL/VAOO wrote: > Thanks Michael & Al! > > I am somewhat of a newbie also with xmlrpc and am having trouble > authenticating to webui. > > I have selinux disabled and have "xmlrpc_rw_enabled: 1" and did a > "service cobblerd restart" after updating to 0.6.2-1 from 0.6.1. I am > unclear what "anyusername = yourpasswordhere" means. It means you can add your own username and password to the config file if you like. The testuser really should be removed (and will be, for when I push -- doing that now). As for your connection problem, I can guess. Since you had a previous cobbler install, check /etc/httpd/conf.d/ and look for a cobbler.conf.rpmnew. That contains the new Apache routing for the XMLRPC parts. Replace the conf.d with the contents of the rpmnew, start httpd, and you should be good to go. I'll add that to the Wiki. > I tried the > testuser with its password but did not have any joy. I also tried my > existing user/password combo. I am assuming auth.conf is similar to > .htaccess but am a little concerned about the clear text aspect. > We're going to SSLify XMLRPC communications later, I'm sure, but here's what we're doing right now. Currently the xmlrpc interface for read-write operations is binding only to localhost. The WebUI is connecting to it via localhost, and you can visit the WebUI with https:// URLs. So your password is never broadcast outside the cobbler server in plaintext. auth.conf is not nearly as fancy as htaccess, it's just a list of username and passwords. And it lives in /etc with permissions such that only root can read it. We'll probably have more sophisticated access control around passwords later (pam maybe?) though for now, attaching only to localhost is secure when the user visits via https://. We should probably put in a redirect to ensure the WebUI isn't visited with http:// just to protect users who don't know better. > Thanks, > Kevin > > >> -----Original Message----- >> From: et-mgmt-tools-bounces at redhat.com >> [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of Michael DeHaan >> Sent: Tuesday, September 25, 2007 6:48 PM >> To: 'et-mgmt-tools at redhat.com' >> Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release >> >> It's new Cobbler & koan time. >> >> The new feature this time is the Cobbler Web UI, for which I have to >> give a big hand to Al Tobey for coding up a *massive* chunk of it. >> The WUI recently gained the ability to edit kickstart files and is >> really shiny (IMHO) at this point. While the rest of cobbler we'll >> consider "stable", the Web UI component should be considered a "beta" >> just for this 1st release. Definitely help test it out though, >> and report back your findings/ideas/comments. If you or >> fellow admins >> need to make tweaks from time to time and need a better >> way of visualizing your provisioning configuration, the WebUI is >> probably for you. >> >> As mentioned before, the setup instructions are on the Wiki, so just >> follow these steps once things are installed: >> >> https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebUi >> >> Unless you've been requesting a feature (in which case, it >> probably was >> added), the other changes are relatively minor >> and you might not notice them. I've sent emails out about >> most of them >> already. One thing admins should note is that >> /var/lib/cobbler/settings is finally packaged as a config file, so >> you'll get an "rpmnew" for the new settings. That will >> be moved into /etc probably in the next release and all will be right >> with config files :) >> >> So ... I'm uploading the 0.6.2 RPM's today for folks to have >> a chance to >> try things out before pushing to the mirrors on Friday. >> >> http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm >> http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm >> >> FYI -- The changelogs might look rather long, though it's >> been about a >> month since the last release and lots of folks have been >> working off the source-control versions. So things are rather well >> tested... >> >> For Cobbler: >> - cobbler repo auto-add to discover yum repos automatically >> (requires a >> rather new yum) >> - fix bug that allows empty mac addresses (None) in dhcpd.conf >> - kickstarts automatically save kickstart file used to >> /root/cobbler.ks >> - allow multiple (comma-seperated) values for --virt-size >> - removed deprecated 'enchant' function (use SSH and koan instead) >> - cleanup of a few unused settings >> - allow for serialization modules to be selected in >> /etc/cobbler/modules.conf >> - patch to allow for reposync of specific repos, even if not >> set to update >> - added --dhcp-tag section for better DHCP customization (esp with >> multiple subnets) >> - added Apache proxying around XMLRPC port for wider network access >> - refactor XMLRPC API and establish a read-write API >> - allow for configuring of read-write XMLRPC users in >> /etc/cobbler/auth.conf >> - WebUI >> - packaged /var/lib/cobbler/settings as a config file >> - added BuildRequires to help build on other platforms >> - relocate cgi-bin files to cgi-bin/cobbler for namespacing >> - fix syslog logging for systems not in the cobbler DB. >> >> For koan >> - Allow for multiple values (comma seperated) to --virt-path >> - Allow for multiple virt file sizes >> - Refactoring around the virt bits >> - Remove legacy sys.path hack as enchant has been removed >> (just use SSH) >> - Try port 80 proxied XMLRPC by default (first) before trying 25151 >> >> As usual, a huge thanks to everyone who submitted patches, emails, or >> hangs out in #cobbler. >> >> --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 Kevin.Maute at WPAFB.AF.MIL Wed Sep 26 15:20:00 2007 From: Kevin.Maute at WPAFB.AF.MIL (Maute, Kevin P CTR AFRL/VAOO) Date: Wed, 26 Sep 2007 11:20:00 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <46FA74F0.9050308@redhat.com> References: <46F9902A.8060205@redhat.com><29A7C5F65D54744CBFD97CD62A33DEE9023848CB@VFOHMLAO13.Enterprise.afmc.ds.af.mil> <46FA74F0.9050308@redhat.com> Message-ID: <29A7C5F65D54744CBFD97CD62A33DEE902384978@VFOHMLAO13.Enterprise.afmc.ds.af.mil> Actually, the only cobbler.conf file in /etc/httpd/conf.d is the old one which is ~ a week old, no *.rpmnew > -----Original Message----- > From: et-mgmt-tools-bounces at redhat.com > [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of Michael DeHaan > Sent: Wednesday, September 26, 2007 11:04 AM > To: Fedora/Linux Management Tools > Subject: Re: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release > > Maute, Kevin P CTR AFRL/VAOO wrote: > > Thanks Michael & Al! > > > > I am somewhat of a newbie also with xmlrpc and am having trouble > > authenticating to webui. > > > > I have selinux disabled and have "xmlrpc_rw_enabled: 1" and did a > > "service cobblerd restart" after updating to 0.6.2-1 from > 0.6.1. I am > > unclear what "anyusername = yourpasswordhere" means. > > It means you can add your own username and password to the > config file > if you like. > The testuser really should be removed (and will be, for when > I push -- > doing that now). > > As for your connection problem, I can guess. > > Since you had a previous cobbler install, check > /etc/httpd/conf.d/ and > look for a cobbler.conf.rpmnew. > That contains the new Apache routing for the XMLRPC parts. > Replace the > conf.d with the contents > of the rpmnew, start httpd, and you should be good to go. > > I'll add that to the Wiki. > > > I tried the > > testuser with its password but did not have any joy. I > also tried my > > existing user/password combo. I am assuming auth.conf is similar to > > .htaccess but am a little concerned about the clear text aspect. > > > > We're going to SSLify XMLRPC communications later, I'm sure, > but here's > what we're doing right now. > > Currently the xmlrpc interface for read-write operations is > binding only > to localhost. The WebUI > is connecting to it via localhost, and you can visit the WebUI with > https:// URLs. So your password > is never broadcast outside the cobbler server in plaintext. > > auth.conf is not nearly as fancy as htaccess, it's just a list of > username and passwords. And it lives in /etc with > permissions such that only root can read it. We'll probably have more > sophisticated access control around > passwords later (pam maybe?) though for now, attaching only > to localhost > is secure when the user visits > via https://. We should probably put in a redirect to ensure > the WebUI > isn't visited with http:// just to protect > users who don't know better. > > > > > > > Thanks, > > Kevin > > > > > >> -----Original Message----- > >> From: et-mgmt-tools-bounces at redhat.com > >> [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of > Michael DeHaan > >> Sent: Tuesday, September 25, 2007 6:48 PM > >> To: 'et-mgmt-tools at redhat.com' > >> Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release > >> > >> It's new Cobbler & koan time. > >> > >> The new feature this time is the Cobbler Web UI, for which > I have to > >> give a big hand to Al Tobey for coding up a *massive* chunk of it. > >> The WUI recently gained the ability to edit kickstart files and is > >> really shiny (IMHO) at this point. While the rest of > cobbler we'll > >> consider "stable", the Web UI component should be > considered a "beta" > >> just for this 1st release. Definitely help test it out though, > >> and report back your findings/ideas/comments. If you or > >> fellow admins > >> need to make tweaks from time to time and need a better > >> way of visualizing your provisioning configuration, the WebUI is > >> probably for you. > >> > >> As mentioned before, the setup instructions are on the > Wiki, so just > >> follow these steps once things are installed: > >> > >> https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebUi > >> > >> Unless you've been requesting a feature (in which case, it > >> probably was > >> added), the other changes are relatively minor > >> and you might not notice them. I've sent emails out about > >> most of them > >> already. One thing admins should note is that > >> /var/lib/cobbler/settings is finally packaged as a config file, so > >> you'll get an "rpmnew" for the new settings. That will > >> be moved into /etc probably in the next release and all > will be right > >> with config files :) > >> > >> So ... I'm uploading the 0.6.2 RPM's today for folks to have > >> a chance to > >> try things out before pushing to the mirrors on Friday. > >> > >> http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm > >> http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm > >> > >> FYI -- The changelogs might look rather long, though it's > >> been about a > >> month since the last release and lots of folks have been > >> working off the source-control versions. So things are > rather well > >> tested... > >> > >> For Cobbler: > >> - cobbler repo auto-add to discover yum repos automatically > >> (requires a > >> rather new yum) > >> - fix bug that allows empty mac addresses (None) in dhcpd.conf > >> - kickstarts automatically save kickstart file used to > >> /root/cobbler.ks > >> - allow multiple (comma-seperated) values for --virt-size > >> - removed deprecated 'enchant' function (use SSH and koan instead) > >> - cleanup of a few unused settings > >> - allow for serialization modules to be selected in > >> /etc/cobbler/modules.conf > >> - patch to allow for reposync of specific repos, even if not > >> set to update > >> - added --dhcp-tag section for better DHCP customization (esp with > >> multiple subnets) > >> - added Apache proxying around XMLRPC port for wider network access > >> - refactor XMLRPC API and establish a read-write API > >> - allow for configuring of read-write XMLRPC users in > >> /etc/cobbler/auth.conf > >> - WebUI > >> - packaged /var/lib/cobbler/settings as a config file > >> - added BuildRequires to help build on other platforms > >> - relocate cgi-bin files to cgi-bin/cobbler for namespacing > >> - fix syslog logging for systems not in the cobbler DB. > >> > >> For koan > >> - Allow for multiple values (comma seperated) to --virt-path > >> - Allow for multiple virt file sizes > >> - Refactoring around the virt bits > >> - Remove legacy sys.path hack as enchant has been removed > >> (just use SSH) > >> - Try port 80 proxied XMLRPC by default (first) before trying 25151 > >> > >> As usual, a huge thanks to everyone who submitted patches, > emails, or > >> hangs out in #cobbler. > >> > >> --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 > > > > _______________________________________________ > 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 Wed Sep 26 15:32:00 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 26 Sep 2007 11:32:00 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <29A7C5F65D54744CBFD97CD62A33DEE902384978@VFOHMLAO13.Enterprise.afmc.ds.af.mil> References: <46F9902A.8060205@redhat.com><29A7C5F65D54744CBFD97CD62A33DEE9023848CB@VFOHMLAO13.Enterprise.afmc.ds.af.mil> <46FA74F0.9050308@redhat.com> <29A7C5F65D54744CBFD97CD62A33DEE902384978@VFOHMLAO13.Enterprise.afmc.ds.af.mil> Message-ID: <46FA7B70.3060207@redhat.com> Maute, Kevin P CTR AFRL/VAOO wrote: > Actually, the only cobbler.conf file in /etc/httpd/conf.d is the old one > which is ~ a week old, no *.rpmnew > If it's a week old, perhaps you already had the right contents, which are... # This configuration file allows cobbler data # to be accessed over HTTP. AliasMatch ^/cblr(.*)?$ "/var/www/cobbler$1" AliasMatch ^/cobbler_track(.*)?$ "/var/www/cobbler$1" AliasMatch ^/cobbler(.*)?$ "/var/www/cobbler$1" Options Indexes FollowSymLinks AllowOverride None Order allow,deny Allow from all ProxyRequests off ProxyPass /cobbler_api http://localhost:25151/ ProxyPassReverse /cobbler_api http://localhost:25151/ # TO DO: SSL ProxyPass /cobbler_api_rw http://localhost:25152/ ProxyPassReverse /cobbler_api_rw http://localhost:25152/ === snip. If you're still not authenticating correctly, I'm not sure what's up. I recently did make a patch that may play a role, however: http://git.fedoraproject.org/?p=hosted/cobbler;a=commitdiff;h=cdc6eef91985545432ec222294b650e0d8269be1 (which will definitely be included in 0.6.2-2 when I push it) You can check /var/log/cobbler/webui.log and /var/log/cobbler/cobblerd.log and you may see some more information there. >> -----Original Message----- >> From: et-mgmt-tools-bounces at redhat.com >> [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of Michael DeHaan >> Sent: Wednesday, September 26, 2007 11:04 AM >> To: Fedora/Linux Management Tools >> Subject: Re: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release >> >> Maute, Kevin P CTR AFRL/VAOO wrote: >> >>> Thanks Michael & Al! >>> >>> I am somewhat of a newbie also with xmlrpc and am having trouble >>> authenticating to webui. >>> >>> I have selinux disabled and have "xmlrpc_rw_enabled: 1" and did a >>> "service cobblerd restart" after updating to 0.6.2-1 from >>> >> 0.6.1. I am >> >>> unclear what "anyusername = yourpasswordhere" means. >>> >> It means you can add your own username and password to the >> config file >> if you like. >> The testuser really should be removed (and will be, for when >> I push -- >> doing that now). >> >> As for your connection problem, I can guess. >> >> Since you had a previous cobbler install, check >> /etc/httpd/conf.d/ and >> look for a cobbler.conf.rpmnew. >> That contains the new Apache routing for the XMLRPC parts. >> Replace the >> conf.d with the contents >> of the rpmnew, start httpd, and you should be good to go. >> >> I'll add that to the Wiki. >> >> >>> I tried the >>> testuser with its password but did not have any joy. I >>> >> also tried my >> >>> existing user/password combo. I am assuming auth.conf is similar to >>> .htaccess but am a little concerned about the clear text aspect. >>> >>> >> We're going to SSLify XMLRPC communications later, I'm sure, >> but here's >> what we're doing right now. >> >> Currently the xmlrpc interface for read-write operations is >> binding only >> to localhost. The WebUI >> is connecting to it via localhost, and you can visit the WebUI with >> https:// URLs. So your password >> is never broadcast outside the cobbler server in plaintext. >> >> auth.conf is not nearly as fancy as htaccess, it's just a list of >> username and passwords. And it lives in /etc with >> permissions such that only root can read it. We'll probably have more >> sophisticated access control around >> passwords later (pam maybe?) though for now, attaching only >> to localhost >> is secure when the user visits >> via https://. We should probably put in a redirect to ensure >> the WebUI >> isn't visited with http:// just to protect >> users who don't know better. >> >> >> >> >> >> >>> Thanks, >>> Kevin >>> >>> >>> >>>> -----Original Message----- >>>> From: et-mgmt-tools-bounces at redhat.com >>>> [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of >>>> >> Michael DeHaan >> >>>> Sent: Tuesday, September 25, 2007 6:48 PM >>>> To: 'et-mgmt-tools at redhat.com' >>>> Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release >>>> >>>> It's new Cobbler & koan time. >>>> >>>> The new feature this time is the Cobbler Web UI, for which >>>> >> I have to >> >>>> give a big hand to Al Tobey for coding up a *massive* chunk of it. >>>> The WUI recently gained the ability to edit kickstart files and is >>>> really shiny (IMHO) at this point. While the rest of >>>> >> cobbler we'll >> >>>> consider "stable", the Web UI component should be >>>> >> considered a "beta" >> >>>> just for this 1st release. Definitely help test it out though, >>>> and report back your findings/ideas/comments. If you or >>>> fellow admins >>>> need to make tweaks from time to time and need a better >>>> way of visualizing your provisioning configuration, the WebUI is >>>> probably for you. >>>> >>>> As mentioned before, the setup instructions are on the >>>> >> Wiki, so just >> >>>> follow these steps once things are installed: >>>> >>>> https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebUi >>>> >>>> Unless you've been requesting a feature (in which case, it >>>> probably was >>>> added), the other changes are relatively minor >>>> and you might not notice them. I've sent emails out about >>>> most of them >>>> already. One thing admins should note is that >>>> /var/lib/cobbler/settings is finally packaged as a config file, so >>>> you'll get an "rpmnew" for the new settings. That will >>>> be moved into /etc probably in the next release and all >>>> >> will be right >> >>>> with config files :) >>>> >>>> So ... I'm uploading the 0.6.2 RPM's today for folks to have >>>> a chance to >>>> try things out before pushing to the mirrors on Friday. >>>> >>>> http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm >>>> http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm >>>> >>>> FYI -- The changelogs might look rather long, though it's >>>> been about a >>>> month since the last release and lots of folks have been >>>> working off the source-control versions. So things are >>>> >> rather well >> >>>> tested... >>>> >>>> For Cobbler: >>>> - cobbler repo auto-add to discover yum repos automatically >>>> (requires a >>>> rather new yum) >>>> - fix bug that allows empty mac addresses (None) in dhcpd.conf >>>> - kickstarts automatically save kickstart file used to >>>> /root/cobbler.ks >>>> - allow multiple (comma-seperated) values for --virt-size >>>> - removed deprecated 'enchant' function (use SSH and koan instead) >>>> - cleanup of a few unused settings >>>> - allow for serialization modules to be selected in >>>> /etc/cobbler/modules.conf >>>> - patch to allow for reposync of specific repos, even if not >>>> set to update >>>> - added --dhcp-tag section for better DHCP customization (esp with >>>> multiple subnets) >>>> - added Apache proxying around XMLRPC port for wider network access >>>> - refactor XMLRPC API and establish a read-write API >>>> - allow for configuring of read-write XMLRPC users in >>>> /etc/cobbler/auth.conf >>>> - WebUI >>>> - packaged /var/lib/cobbler/settings as a config file >>>> - added BuildRequires to help build on other platforms >>>> - relocate cgi-bin files to cgi-bin/cobbler for namespacing >>>> - fix syslog logging for systems not in the cobbler DB. >>>> >>>> For koan >>>> - Allow for multiple values (comma seperated) to --virt-path >>>> - Allow for multiple virt file sizes >>>> - Refactoring around the virt bits >>>> - Remove legacy sys.path hack as enchant has been removed >>>> (just use SSH) >>>> - Try port 80 proxied XMLRPC by default (first) before trying 25151 >>>> >>>> As usual, a huge thanks to everyone who submitted patches, >>>> >> emails, or >> >>>> hangs out in #cobbler. >>>> >>>> --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 >>> >>> >> _______________________________________________ >> 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 Kevin.Maute at WPAFB.AF.MIL Wed Sep 26 15:46:37 2007 From: Kevin.Maute at WPAFB.AF.MIL (Maute, Kevin P CTR AFRL/VAOO) Date: Wed, 26 Sep 2007 11:46:37 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <46FA7B70.3060207@redhat.com> References: <46F9902A.8060205@redhat.com><29A7C5F65D54744CBFD97CD62A33DEE9023848CB@VFOHMLAO13.Enterprise.afmc.ds.af.mil> <46FA74F0.9050308@redhat.com><29A7C5F65D54744CBFD97CD62A33DEE902384978@VFOHMLAO13.Enterprise.afmc.ds.af.mil> <46FA7B70.3060207@redhat.com> Message-ID: <29A7C5F65D54744CBFD97CD62A33DEE9023849CE@VFOHMLAO13.Enterprise.afmc.ds.af.mil> Yes, they are identical... I don't think I ever looked at this file during my ramp up to know what the old one had. I also tried this with and without iptables running > -----Original Message----- > From: et-mgmt-tools-bounces at redhat.com > [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of Michael DeHaan > Sent: Wednesday, September 26, 2007 11:32 AM > To: Fedora/Linux Management Tools > Subject: Re: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release > > Maute, Kevin P CTR AFRL/VAOO wrote: > > Actually, the only cobbler.conf file in /etc/httpd/conf.d > is the old one > > which is ~ a week old, no *.rpmnew > > > If it's a week old, perhaps you already had the right > contents, which are... > > # This configuration file allows cobbler data > # to be accessed over HTTP. > > AliasMatch ^/cblr(.*)?$ "/var/www/cobbler$1" > AliasMatch ^/cobbler_track(.*)?$ "/var/www/cobbler$1" > AliasMatch ^/cobbler(.*)?$ "/var/www/cobbler$1" > > > Options Indexes FollowSymLinks > AllowOverride None > Order allow,deny > Allow from all > > > ProxyRequests off > > ProxyPass /cobbler_api http://localhost:25151/ > ProxyPassReverse /cobbler_api http://localhost:25151/ > > # TO DO: SSL > > ProxyPass /cobbler_api_rw http://localhost:25152/ > ProxyPassReverse /cobbler_api_rw http://localhost:25152/ > > > === > > snip. > > If you're still not authenticating correctly, I'm not sure what's up. > I recently did make a patch that may play a role, however: > http://git.fedoraproject.org/?p=hosted/cobbler;a=commitdiff;h= cdc6eef91985545432ec222294b650e0d8269be1 > (which will definitely be included in 0.6.2-2 when I push it) > > You can check /var/log/cobbler/webui.log and > /var/log/cobbler/cobblerd.log and you may see some more > information there. > > >> -----Original Message----- > >> From: et-mgmt-tools-bounces at redhat.com > >> [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of > Michael DeHaan > >> Sent: Wednesday, September 26, 2007 11:04 AM > >> To: Fedora/Linux Management Tools > >> Subject: Re: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release > >> > >> Maute, Kevin P CTR AFRL/VAOO wrote: > >> > >>> Thanks Michael & Al! > >>> > >>> I am somewhat of a newbie also with xmlrpc and am having trouble > >>> authenticating to webui. > >>> > >>> I have selinux disabled and have "xmlrpc_rw_enabled: 1" and did a > >>> "service cobblerd restart" after updating to 0.6.2-1 from > >>> > >> 0.6.1. I am > >> > >>> unclear what "anyusername = yourpasswordhere" means. > >>> > >> It means you can add your own username and password to the > >> config file > >> if you like. > >> The testuser really should be removed (and will be, for when > >> I push -- > >> doing that now). > >> > >> As for your connection problem, I can guess. > >> > >> Since you had a previous cobbler install, check > >> /etc/httpd/conf.d/ and > >> look for a cobbler.conf.rpmnew. > >> That contains the new Apache routing for the XMLRPC parts. > >> Replace the > >> conf.d with the contents > >> of the rpmnew, start httpd, and you should be good to go. > >> > >> I'll add that to the Wiki. > >> > >> > >>> I tried the > >>> testuser with its password but did not have any joy. I > >>> > >> also tried my > >> > >>> existing user/password combo. I am assuming auth.conf is > similar to > >>> .htaccess but am a little concerned about the clear text aspect. > >>> > >>> > >> We're going to SSLify XMLRPC communications later, I'm sure, > >> but here's > >> what we're doing right now. > >> > >> Currently the xmlrpc interface for read-write operations is > >> binding only > >> to localhost. The WebUI > >> is connecting to it via localhost, and you can visit the > WebUI with > >> https:// URLs. So your password > >> is never broadcast outside the cobbler server in plaintext. > >> > >> auth.conf is not nearly as fancy as htaccess, it's just a list of > >> username and passwords. And it lives in /etc with > >> permissions such that only root can read it. We'll > probably have more > >> sophisticated access control around > >> passwords later (pam maybe?) though for now, attaching only > >> to localhost > >> is secure when the user visits > >> via https://. We should probably put in a redirect to ensure > >> the WebUI > >> isn't visited with http:// just to protect > >> users who don't know better. > >> > >> > >> > >> > >> > >> > >>> Thanks, > >>> Kevin > >>> > >>> > >>> > >>>> -----Original Message----- > >>>> From: et-mgmt-tools-bounces at redhat.com > >>>> [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of > >>>> > >> Michael DeHaan > >> > >>>> Sent: Tuesday, September 25, 2007 6:48 PM > >>>> To: 'et-mgmt-tools at redhat.com' > >>>> Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release > >>>> > >>>> It's new Cobbler & koan time. > >>>> > >>>> The new feature this time is the Cobbler Web UI, for which > >>>> > >> I have to > >> > >>>> give a big hand to Al Tobey for coding up a *massive* > chunk of it. > >>>> The WUI recently gained the ability to edit kickstart > files and is > >>>> really shiny (IMHO) at this point. While the rest of > >>>> > >> cobbler we'll > >> > >>>> consider "stable", the Web UI component should be > >>>> > >> considered a "beta" > >> > >>>> just for this 1st release. Definitely help test it out though, > >>>> and report back your findings/ideas/comments. If you or > >>>> fellow admins > >>>> need to make tweaks from time to time and need a better > >>>> way of visualizing your provisioning configuration, the WebUI is > >>>> probably for you. > >>>> > >>>> As mentioned before, the setup instructions are on the > >>>> > >> Wiki, so just > >> > >>>> follow these steps once things are installed: > >>>> > >>>> > https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebUi > >>>> > >>>> Unless you've been requesting a feature (in which case, it > >>>> probably was > >>>> added), the other changes are relatively minor > >>>> and you might not notice them. I've sent emails out about > >>>> most of them > >>>> already. One thing admins should note is that > >>>> /var/lib/cobbler/settings is finally packaged as a > config file, so > >>>> you'll get an "rpmnew" for the new settings. That will > >>>> be moved into /etc probably in the next release and all > >>>> > >> will be right > >> > >>>> with config files :) > >>>> > >>>> So ... I'm uploading the 0.6.2 RPM's today for folks to have > >>>> a chance to > >>>> try things out before pushing to the mirrors on Friday. > >>>> > >>>> http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm > >>>> http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm > >>>> > >>>> FYI -- The changelogs might look rather long, though it's > >>>> been about a > >>>> month since the last release and lots of folks have been > >>>> working off the source-control versions. So things are > >>>> > >> rather well > >> > >>>> tested... > >>>> > >>>> For Cobbler: > >>>> - cobbler repo auto-add to discover yum repos automatically > >>>> (requires a > >>>> rather new yum) > >>>> - fix bug that allows empty mac addresses (None) in dhcpd.conf > >>>> - kickstarts automatically save kickstart file used to > >>>> /root/cobbler.ks > >>>> - allow multiple (comma-seperated) values for --virt-size > >>>> - removed deprecated 'enchant' function (use SSH and > koan instead) > >>>> - cleanup of a few unused settings > >>>> - allow for serialization modules to be selected in > >>>> /etc/cobbler/modules.conf > >>>> - patch to allow for reposync of specific repos, even if not > >>>> set to update > >>>> - added --dhcp-tag section for better DHCP customization > (esp with > >>>> multiple subnets) > >>>> - added Apache proxying around XMLRPC port for wider > network access > >>>> - refactor XMLRPC API and establish a read-write API > >>>> - allow for configuring of read-write XMLRPC users in > >>>> /etc/cobbler/auth.conf > >>>> - WebUI > >>>> - packaged /var/lib/cobbler/settings as a config file > >>>> - added BuildRequires to help build on other platforms > >>>> - relocate cgi-bin files to cgi-bin/cobbler for namespacing > >>>> - fix syslog logging for systems not in the cobbler DB. > >>>> > >>>> For koan > >>>> - Allow for multiple values (comma seperated) to --virt-path > >>>> - Allow for multiple virt file sizes > >>>> - Refactoring around the virt bits > >>>> - Remove legacy sys.path hack as enchant has been removed > >>>> (just use SSH) > >>>> - Try port 80 proxied XMLRPC by default (first) before > trying 25151 > >>>> > >>>> As usual, a huge thanks to everyone who submitted patches, > >>>> > >> emails, or > >> > >>>> hangs out in #cobbler. > >>>> > >>>> --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 > >>> > >>> > >> _______________________________________________ > >> 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 Kevin.Maute at WPAFB.AF.MIL Wed Sep 26 15:51:31 2007 From: Kevin.Maute at WPAFB.AF.MIL (Maute, Kevin P CTR AFRL/VAOO) Date: Wed, 26 Sep 2007 11:51:31 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <29A7C5F65D54744CBFD97CD62A33DEE9023849CE@VFOHMLAO13.Enterprise.afmc.ds.af.mil> References: <46F9902A.8060205@redhat.com><29A7C5F65D54744CBFD97CD62A33DEE9023848CB@VFOHMLAO13.Enterprise.afmc.ds.af.mil> <46FA74F0.9050308@redhat.com><29A7C5F65D54744CBFD97CD62A33DEE902384978@VFOHMLAO13.Enterprise.afmc.ds.af.mil><46FA7B70.3060207@redhat.com> <29A7C5F65D54744CBFD97CD62A33DEE9023849CE@VFOHMLAO13.Enterprise.afmc.ds.af.mil> Message-ID: <29A7C5F65D54744CBFD97CD62A33DEE9023849E3@VFOHMLAO13.Enterprise.afmc.ds.af.mil> DOH! You might want to mention what you told me earlier about restarting httpd in the wiki also for those of use who tend to more literal than they should be!!! > -----Original Message----- > From: et-mgmt-tools-bounces at redhat.com > [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of Maute, > Kevin P CTR AFRL/VAOO > Sent: Wednesday, September 26, 2007 11:47 AM > To: Fedora/Linux Management Tools > Subject: RE: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release > > Yes, they are identical... I don't think I ever looked at this file > during my ramp up to know what the old one had. I also tried > this with > and without iptables running > > > -----Original Message----- > > From: et-mgmt-tools-bounces at redhat.com > > [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of > Michael DeHaan > > Sent: Wednesday, September 26, 2007 11:32 AM > > To: Fedora/Linux Management Tools > > Subject: Re: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release > > > > Maute, Kevin P CTR AFRL/VAOO wrote: > > > Actually, the only cobbler.conf file in /etc/httpd/conf.d > > is the old one > > > which is ~ a week old, no *.rpmnew > > > > > If it's a week old, perhaps you already had the right > > contents, which are... > > > > # This configuration file allows cobbler data > > # to be accessed over HTTP. > > > > AliasMatch ^/cblr(.*)?$ "/var/www/cobbler$1" > > AliasMatch ^/cobbler_track(.*)?$ "/var/www/cobbler$1" > > AliasMatch ^/cobbler(.*)?$ "/var/www/cobbler$1" > > > > > > Options Indexes FollowSymLinks > > AllowOverride None > > Order allow,deny > > Allow from all > > > > > > ProxyRequests off > > > > ProxyPass /cobbler_api http://localhost:25151/ > > ProxyPassReverse /cobbler_api http://localhost:25151/ > > > > # TO DO: SSL > > > > ProxyPass /cobbler_api_rw http://localhost:25152/ > > ProxyPassReverse /cobbler_api_rw http://localhost:25152/ > > > > > > === > > > > snip. > > > > If you're still not authenticating correctly, I'm not sure > what's up. > > I recently did make a patch that may play a role, however: > > http://git.fedoraproject.org/?p=hosted/cobbler;a=commitdiff;h= > cdc6eef91985545432ec222294b650e0d8269be1 > > (which will definitely be included in 0.6.2-2 when I push it) > > > > You can check /var/log/cobbler/webui.log and > > /var/log/cobbler/cobblerd.log and you may see some more > > information there. > > > > >> -----Original Message----- > > >> From: et-mgmt-tools-bounces at redhat.com > > >> [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of > > Michael DeHaan > > >> Sent: Wednesday, September 26, 2007 11:04 AM > > >> To: Fedora/Linux Management Tools > > >> Subject: Re: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 > (pre) release > > >> > > >> Maute, Kevin P CTR AFRL/VAOO wrote: > > >> > > >>> Thanks Michael & Al! > > >>> > > >>> I am somewhat of a newbie also with xmlrpc and am having trouble > > >>> authenticating to webui. > > >>> > > >>> I have selinux disabled and have "xmlrpc_rw_enabled: 1" > and did a > > >>> "service cobblerd restart" after updating to 0.6.2-1 from > > >>> > > >> 0.6.1. I am > > >> > > >>> unclear what "anyusername = yourpasswordhere" means. > > >>> > > >> It means you can add your own username and password to the > > >> config file > > >> if you like. > > >> The testuser really should be removed (and will be, for when > > >> I push -- > > >> doing that now). > > >> > > >> As for your connection problem, I can guess. > > >> > > >> Since you had a previous cobbler install, check > > >> /etc/httpd/conf.d/ and > > >> look for a cobbler.conf.rpmnew. > > >> That contains the new Apache routing for the XMLRPC parts. > > >> Replace the > > >> conf.d with the contents > > >> of the rpmnew, start httpd, and you should be good to go. > > >> > > >> I'll add that to the Wiki. > > >> > > >> > > >>> I tried the > > >>> testuser with its password but did not have any joy. I > > >>> > > >> also tried my > > >> > > >>> existing user/password combo. I am assuming auth.conf is > > similar to > > >>> .htaccess but am a little concerned about the clear text aspect. > > >>> > > >>> > > >> We're going to SSLify XMLRPC communications later, I'm sure, > > >> but here's > > >> what we're doing right now. > > >> > > >> Currently the xmlrpc interface for read-write operations is > > >> binding only > > >> to localhost. The WebUI > > >> is connecting to it via localhost, and you can visit the > > WebUI with > > >> https:// URLs. So your password > > >> is never broadcast outside the cobbler server in plaintext. > > >> > > >> auth.conf is not nearly as fancy as htaccess, it's just > a list of > > >> username and passwords. And it lives in /etc with > > >> permissions such that only root can read it. We'll > > probably have more > > >> sophisticated access control around > > >> passwords later (pam maybe?) though for now, attaching only > > >> to localhost > > >> is secure when the user visits > > >> via https://. We should probably put in a redirect to ensure > > >> the WebUI > > >> isn't visited with http:// just to protect > > >> users who don't know better. > > >> > > >> > > >> > > >> > > >> > > >> > > >>> Thanks, > > >>> Kevin > > >>> > > >>> > > >>> > > >>>> -----Original Message----- > > >>>> From: et-mgmt-tools-bounces at redhat.com > > >>>> [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of > > >>>> > > >> Michael DeHaan > > >> > > >>>> Sent: Tuesday, September 25, 2007 6:48 PM > > >>>> To: 'et-mgmt-tools at redhat.com' > > >>>> Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release > > >>>> > > >>>> It's new Cobbler & koan time. > > >>>> > > >>>> The new feature this time is the Cobbler Web UI, for which > > >>>> > > >> I have to > > >> > > >>>> give a big hand to Al Tobey for coding up a *massive* > > chunk of it. > > >>>> The WUI recently gained the ability to edit kickstart > > files and is > > >>>> really shiny (IMHO) at this point. While the rest of > > >>>> > > >> cobbler we'll > > >> > > >>>> consider "stable", the Web UI component should be > > >>>> > > >> considered a "beta" > > >> > > >>>> just for this 1st release. Definitely help test it out though, > > >>>> and report back your findings/ideas/comments. If you or > > >>>> fellow admins > > >>>> need to make tweaks from time to time and need a better > > >>>> way of visualizing your provisioning configuration, > the WebUI is > > >>>> probably for you. > > >>>> > > >>>> As mentioned before, the setup instructions are on the > > >>>> > > >> Wiki, so just > > >> > > >>>> follow these steps once things are installed: > > >>>> > > >>>> > > https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebUi > > >>>> > > >>>> Unless you've been requesting a feature (in which case, it > > >>>> probably was > > >>>> added), the other changes are relatively minor > > >>>> and you might not notice them. I've sent emails out about > > >>>> most of them > > >>>> already. One thing admins should note is that > > >>>> /var/lib/cobbler/settings is finally packaged as a > > config file, so > > >>>> you'll get an "rpmnew" for the new settings. That will > > >>>> be moved into /etc probably in the next release and all > > >>>> > > >> will be right > > >> > > >>>> with config files :) > > >>>> > > >>>> So ... I'm uploading the 0.6.2 RPM's today for folks to have > > >>>> a chance to > > >>>> try things out before pushing to the mirrors on Friday. > > >>>> > > >>>> http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm > > >>>> http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm > > >>>> > > >>>> FYI -- The changelogs might look rather long, though it's > > >>>> been about a > > >>>> month since the last release and lots of folks have been > > >>>> working off the source-control versions. So things are > > >>>> > > >> rather well > > >> > > >>>> tested... > > >>>> > > >>>> For Cobbler: > > >>>> - cobbler repo auto-add to discover yum repos automatically > > >>>> (requires a > > >>>> rather new yum) > > >>>> - fix bug that allows empty mac addresses (None) in dhcpd.conf > > >>>> - kickstarts automatically save kickstart file used to > > >>>> /root/cobbler.ks > > >>>> - allow multiple (comma-seperated) values for --virt-size > > >>>> - removed deprecated 'enchant' function (use SSH and > > koan instead) > > >>>> - cleanup of a few unused settings > > >>>> - allow for serialization modules to be selected in > > >>>> /etc/cobbler/modules.conf > > >>>> - patch to allow for reposync of specific repos, even if not > > >>>> set to update > > >>>> - added --dhcp-tag section for better DHCP customization > > (esp with > > >>>> multiple subnets) > > >>>> - added Apache proxying around XMLRPC port for wider > > network access > > >>>> - refactor XMLRPC API and establish a read-write API > > >>>> - allow for configuring of read-write XMLRPC users in > > >>>> /etc/cobbler/auth.conf > > >>>> - WebUI > > >>>> - packaged /var/lib/cobbler/settings as a config file > > >>>> - added BuildRequires to help build on other platforms > > >>>> - relocate cgi-bin files to cgi-bin/cobbler for namespacing > > >>>> - fix syslog logging for systems not in the cobbler DB. > > >>>> > > >>>> For koan > > >>>> - Allow for multiple values (comma seperated) to --virt-path > > >>>> - Allow for multiple virt file sizes > > >>>> - Refactoring around the virt bits > > >>>> - Remove legacy sys.path hack as enchant has been removed > > >>>> (just use SSH) > > >>>> - Try port 80 proxied XMLRPC by default (first) before > > trying 25151 > > >>>> > > >>>> As usual, a huge thanks to everyone who submitted patches, > > >>>> > > >> emails, or > > >> > > >>>> hangs out in #cobbler. > > >>>> > > >>>> --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 > > >>> > > >>> > > >> _______________________________________________ > > >> 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 > > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From Daniel.Hennessey at aah.co.uk Wed Sep 26 16:22:18 2007 From: Daniel.Hennessey at aah.co.uk (Hennessey Daniel) Date: Wed, 26 Sep 2007 17:22:18 +0100 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release Message-ID: <03BB5BD4CF43594B9542DCCD4E7989900B046CE7@GBW607SC0054.GB-WS.net> Hey! I have upgraded my cobbler installation but I seem to be having some selinux problems when trying to use the shiny, new webUI. It appears that the cobbler CGI script which runs as context "httpd_sys_script_t" is trying to write to the cobbler logfile "/var/log/cobbler/cobblerd.log" which has a context of "var_log_t". Can anybody tell me how to get around this problem (preferably without disabling selinux)? Cheers in advance. Dan -----Original Message----- From: et-mgmt-tools-bounces at redhat.com [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of Michael DeHaan Sent: 25 September 2007 23:48 To: 'et-mgmt-tools at redhat.com' Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release It's new Cobbler & koan time. The new feature this time is the Cobbler Web UI, for which I have to give a big hand to Al Tobey for coding up a *massive* chunk of it. The WUI recently gained the ability to edit kickstart files and is really shiny (IMHO) at this point. While the rest of cobbler we'll consider "stable", the Web UI component should be considered a "beta" just for this 1st release. Definitely help test it out though, and report back your findings/ideas/comments. If you or fellow admins need to make tweaks from time to time and need a better way of visualizing your provisioning configuration, the WebUI is probably for you. As mentioned before, the setup instructions are on the Wiki, so just follow these steps once things are installed: https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebUi Unless you've been requesting a feature (in which case, it probably was added), the other changes are relatively minor and you might not notice them. I've sent emails out about most of them already. One thing admins should note is that /var/lib/cobbler/settings is finally packaged as a config file, so you'll get an "rpmnew" for the new settings. That will be moved into /etc probably in the next release and all will be right with config files :) So ... I'm uploading the 0.6.2 RPM's today for folks to have a chance to try things out before pushing to the mirrors on Friday. http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm FYI -- The changelogs might look rather long, though it's been about a month since the last release and lots of folks have been working off the source-control versions. So things are rather well tested... For Cobbler: - cobbler repo auto-add to discover yum repos automatically (requires a rather new yum) - fix bug that allows empty mac addresses (None) in dhcpd.conf - kickstarts automatically save kickstart file used to /root/cobbler.ks - allow multiple (comma-seperated) values for --virt-size - removed deprecated 'enchant' function (use SSH and koan instead) - cleanup of a few unused settings - allow for serialization modules to be selected in /etc/cobbler/modules.conf - patch to allow for reposync of specific repos, even if not set to update - added --dhcp-tag section for better DHCP customization (esp with multiple subnets) - added Apache proxying around XMLRPC port for wider network access - refactor XMLRPC API and establish a read-write API - allow for configuring of read-write XMLRPC users in /etc/cobbler/auth.conf - WebUI - packaged /var/lib/cobbler/settings as a config file - added BuildRequires to help build on other platforms - relocate cgi-bin files to cgi-bin/cobbler for namespacing - fix syslog logging for systems not in the cobbler DB. For koan - Allow for multiple values (comma seperated) to --virt-path - Allow for multiple virt file sizes - Refactoring around the virt bits - Remove legacy sys.path hack as enchant has been removed (just use SSH) - Try port 80 proxied XMLRPC by default (first) before trying 25151 As usual, a huge thanks to everyone who submitted patches, emails, or hangs out in #cobbler. --Michael _______________________________________________ et-mgmt-tools mailing list et-mgmt-tools at redhat.com https://www.redhat.com/mailman/listinfo/et-mgmt-tools ************************************************************************ DISCLAIMER The information contained in this e-mail is confidential and is intended for the recipient only. If you have received it in error, please notify us immediately by reply e-mail and then delete it from your system. Please do not copy it or use it for any other purposes, or disclose the content of the e-mail to any other person or store or copy the information in any medium. The views contained in this e-mail are those of the author and not necessarily those of Admenta UK Group. Admenta UK plc is a company incorporated in England and Wales under company number 3011757 and whose registered office is at Sapphire Court, Walsgrave Triangle, Coventry, CV2 2TX ************************************************************************ From mdehaan at redhat.com Wed Sep 26 16:30:07 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 26 Sep 2007 12:30:07 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <03BB5BD4CF43594B9542DCCD4E7989900B046CE7@GBW607SC0054.GB-WS.net> References: <03BB5BD4CF43594B9542DCCD4E7989900B046CE7@GBW607SC0054.GB-WS.net> Message-ID: <46FA890F.4020501@redhat.com> Hennessey Daniel wrote: > Hey! > > I have upgraded my cobbler installation but I seem to be having some > selinux problems when trying to use the shiny, new webUI. It appears > that the cobbler CGI script which runs as context "httpd_sys_script_t" > is trying to write to the cobbler logfile > "/var/log/cobbler/cobblerd.log" which has a context of "var_log_t". > > Can anybody tell me how to get around this problem (preferably without > disabling selinux)? > > Cheers in advance. > > Dan > It shouldn't be writting to that log file, but instead /var/log/cobbler/wui.log. That's probably where you are seeing the SELinux error. I've been told I should move the logfile to /var/log/httpd/cobbler_wui.log and that seems like it will solve the problem. Until then, you could try running audit2allow or the equivalent... --Michael > -----Original Message----- > From: et-mgmt-tools-bounces at redhat.com > [mailto:et-mgmt-tools-bounces at redhat.com] On Behalf Of Michael DeHaan > Sent: 25 September 2007 23:48 > To: 'et-mgmt-tools at redhat.com' > Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release > > > It's new Cobbler & koan time. > > The new feature this time is the Cobbler Web UI, for which I have to > give a big hand to Al Tobey for coding up a *massive* chunk of it. The > WUI recently gained the ability to edit kickstart files and is > really shiny (IMHO) at this point. While the rest of cobbler we'll > consider "stable", the Web UI component should be considered a "beta" > just for this 1st release. Definitely help test it out though, > and report back your findings/ideas/comments. If you or fellow admins > need to make tweaks from time to time and need a better > way of visualizing your provisioning configuration, the WebUI is > probably for you. > > As mentioned before, the setup instructions are on the Wiki, so just > follow these steps once things are installed: > > https://hosted.fedoraproject.org/projects/cobbler/wiki/CobblerWebUi > > Unless you've been requesting a feature (in which case, it probably was > added), the other changes are relatively minor > and you might not notice them. I've sent emails out about most of them > already. One thing admins should note is that > /var/lib/cobbler/settings is finally packaged as a config file, so > you'll get an "rpmnew" for the new settings. That will > be moved into /etc probably in the next release and all will be right > with config files :) > > So ... I'm uploading the 0.6.2 RPM's today for folks to have a chance to > > try things out before pushing to the mirrors on Friday. > > http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm > http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm > > FYI -- The changelogs might look rather long, though it's been about a > month since the last release and lots of folks have been > working off the source-control versions. So things are rather well > tested... > > For Cobbler: > - cobbler repo auto-add to discover yum repos automatically (requires a > > rather new yum) > - fix bug that allows empty mac addresses (None) in dhcpd.conf > - kickstarts automatically save kickstart file used to /root/cobbler.ks > - allow multiple (comma-seperated) values for --virt-size > - removed deprecated 'enchant' function (use SSH and koan instead) > - cleanup of a few unused settings > - allow for serialization modules to be selected in > /etc/cobbler/modules.conf > - patch to allow for reposync of specific repos, even if not set to > update > - added --dhcp-tag section for better DHCP customization (esp with > multiple subnets) > - added Apache proxying around XMLRPC port for wider network access > - refactor XMLRPC API and establish a read-write API > - allow for configuring of read-write XMLRPC users in > /etc/cobbler/auth.conf > - WebUI > - packaged /var/lib/cobbler/settings as a config file > - added BuildRequires to help build on other platforms > - relocate cgi-bin files to cgi-bin/cobbler for namespacing > - fix syslog logging for systems not in the cobbler DB. > > For koan > - Allow for multiple values (comma seperated) to --virt-path > - Allow for multiple virt file sizes > - Refactoring around the virt bits > - Remove legacy sys.path hack as enchant has been removed (just use SSH) > - Try port 80 proxied XMLRPC by default (first) before trying 25151 > > As usual, a huge thanks to everyone who submitted patches, emails, or > hangs out in #cobbler. > > --Michael > > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > ************************************************************************ > DISCLAIMER > The information contained in this e-mail is confidential and is intended > for the recipient only. > If you have received it in error, please notify us immediately by reply > e-mail and then delete it from your system. Please do not copy it or > use it for any other purposes, or disclose the content of the e-mail > to any other person or store or copy the information in any medium. > The views contained in this e-mail are those of the author and not > necessarily those of Admenta UK Group. > Admenta UK plc is a company incorporated in England and Wales > under company number 3011757 and whose registered office > is at Sapphire Court, Walsgrave Triangle, Coventry, CV2 2TX > ************************************************************************ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Wed Sep 26 16:41:42 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Wed, 26 Sep 2007 18:41:42 +0200 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <46F9902A.8060205@redhat.com> References: <46F9902A.8060205@redhat.com> Message-ID: <20070926184142.24b3c639@python3.es.egwn.lan> Michael DeHaan wrote : > The new feature this time is the Cobbler Web UI, for which I have to > give a big hand to Al Tobey for coding up a *massive* chunk of it. > The WUI recently gained the ability to edit kickstart files and is > really shiny (IMHO) at this point. While the rest of cobbler we'll > consider "stable", the Web UI component should be considered a "beta" > just for this 1st release. Definitely help test it out though, > and report back your findings/ideas/comments. If you or fellow admins > need to make tweaks from time to time and need a better > way of visualizing your provisioning configuration, the WebUI is > probably for you. The WebUI is really nice. I just got this error, though, when I removed all profiles and systems, then clicked on SYSTEM (under ADD). Maybe because I have no profiles available yet ;-) All other items in the menu work fine so far. [lots of purple and pink python errors] ParseError: Invalid end directive Line 106, column 89 in file /usr/share/cobbler/webui_templates/system_edit.tmpl Line|Cheetah Code ----|------------------------------------------------------------- 103 | #end if 104 | /> 105 | #if not $system 106 | random# #end ^ 107 |

Example: AA:BB:CC:DD:EE:FF

108 | 109 | col = None extMsg = '' lineno = None msg = 'Invalid end directive' report = > stream = Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.14 0.36 1.15 From mdehaan at redhat.com Wed Sep 26 16:41:25 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 26 Sep 2007 12:41:25 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <20070926184142.24b3c639@python3.es.egwn.lan> References: <46F9902A.8060205@redhat.com> <20070926184142.24b3c639@python3.es.egwn.lan> Message-ID: <46FA8BB5.9050006@redhat.com> Matthias Saou wrote: > Michael DeHaan wrote : > > >> The new feature this time is the Cobbler Web UI, for which I have to >> give a big hand to Al Tobey for coding up a *massive* chunk of it. >> The WUI recently gained the ability to edit kickstart files and is >> really shiny (IMHO) at this point. While the rest of cobbler we'll >> consider "stable", the Web UI component should be considered a "beta" >> just for this 1st release. Definitely help test it out though, >> and report back your findings/ideas/comments. If you or fellow admins >> need to make tweaks from time to time and need a better >> way of visualizing your provisioning configuration, the WebUI is >> probably for you. >> > > The WebUI is really nice. I just got this error, though, when I > removed all profiles and systems, then clicked on SYSTEM (under ADD). > Maybe because I have no profiles available yet ;-) > > All other items in the menu work fine so far. > > [lots of purple and pink python errors] > > ParseError: Invalid end directive Line 106, column 89 in > file /usr/share/cobbler/webui_templates/system_edit.tmpl Line|Cheetah > Code ----|------------------------------------------------------------- > 103 | #end if 104 | /> 105 | #if not $system 106 | random# #end ^ 107 |

class="context-tip">Example: AA:BB:CC:DD:EE:FF

108 | 109 | > col = None extMsg = '' lineno = None msg = 'Invalid end > directive' report = > stream = > > > Matthias > > Ok, I've pushed a fix. Thanks! git clone git://git.fedoraproject.org/git/hosted/cobbler From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Wed Sep 26 16:58:43 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Wed, 26 Sep 2007 18:58:43 +0200 Subject: [et-mgmt-tools] Cobbler and Koan for _my_ needs Message-ID: <20070926185843.5f040191@python3.es.egwn.lan> Hi, I hope this won't seem like a selfish request, I'll just try to explain how cobbler and koan could much better fit my needs, and possibly those of others! My setup : I run Xen PV guests. LOTS of them. What I do to ease their management is use unpartitioned LVM logical volumes from the host as block devices, one for "/" and the other for swap. I don't let any tools create the guest config files, I create them myself "by hand" once, then deploy them through an rpm package to all the Xen hosts I have available. This way I have a central location to check for all guest configuration and MAC addresses. The initial guest install is where I'd like to use cobbler and koan in order to save time. I currently extract and modify a pre-existing RHEL5 install, but that's another story ;-) Problem : Koan wants to create the Xen guest configuration files for me. I'd like it to be able to use an existing file and have koan extract the settings it needs from there. Here is an example Xen guest config file I use : --- # Xen Configuration File name = "cobbler" memory = "512" maxmem = "512" disk = [ 'phy:/dev/data/cobbler,xvda,w', 'phy:/dev/data/swapcobbler,xvdb,w' ] vif = [ 'mac=00:16:3e:5f:76:6e, bridge=xenbr0', 'mac=00:16:3e:5f:76:6f, bridge=xenbr1' ] uuid = '5f766eef-8cf3-c287-1635-115e103561d0' bootloader = '/usr/bin/pygrub' vcpus = 2 on_reboot = 'restart' on_crash = 'restart' --- I generate the uuid myself (md5sum of the full host name), then use the 6 first characters in the eth0 mac address, increment it for the eth1 mac address. Ideally, I'd like to be able to tell koan to use this existing "cobbler" file and use xenbr1 for everything it needs to do (since dhcpd and cobbler are only listening on that "private" network), then wait, and get a working "cobbler" system up and running :-) Is this something that could be achieved? Am I the only one to have multiple block devices, multiple network interfaces, and like having fixed MAC addresses for my Xen guests? Currently, koan might be useful to me, but once I get the guest running, I'd still need to move stuff around block devices and change bits and pieces of configuration to get things into place as I want them. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.42 0.44 0.97 From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Wed Sep 26 17:06:46 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Wed, 26 Sep 2007 19:06:46 +0200 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <46FA8BB5.9050006@redhat.com> References: <46F9902A.8060205@redhat.com> <20070926184142.24b3c639@python3.es.egwn.lan> <46FA8BB5.9050006@redhat.com> Message-ID: <20070926190646.44fda0d0@python3.es.egwn.lan> Michael DeHaan wrote : > Ok, I've pushed a fix. Thanks! > > git clone git://git.fedoraproject.org/git/hosted/cobbler Oh, I've just diffed git against the 0.6.2 snapshot : - Having added "/var/log/httpd/cobbler_wui.log" to cobbler/config/cobblerd_rotate is wrong, since the httpd logrotate.d file already has "/var/log/httpd/*log", so logrotate will now report a duplicate entry for that file. It should be removed. - Having changed this is also wrong : - /etc/init.d/cobblerd restart + /etc/init.d/cobblerd condrestart As the /var/lock/subsys/cobblerd test is still present and if cobblerd is running, root will get an email with the "stopping cobblerd... OK, starting cobblerd... OK" lines. You should probably use : /sbin/service cobblerd condrestart >/dev/null Adn remove the subsys test, which will output nothing if cobblerd isn't running, nothing if it is because of the redirection, and output any possible errors (which shouldn't happen, but... you know ;-)) Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.47 0.37 0.51 From mdehaan at redhat.com Wed Sep 26 17:41:25 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 26 Sep 2007 13:41:25 -0400 Subject: [et-mgmt-tools] Cobbler and Koan for _my_ needs In-Reply-To: <20070926185843.5f040191@python3.es.egwn.lan> References: <20070926185843.5f040191@python3.es.egwn.lan> Message-ID: <46FA99C5.6010806@redhat.com> Matthias Saou wrote: > Hi, > > I hope this won't seem like a selfish request, I'll just try to explain > how cobbler and koan could much better fit my needs, and possibly those > of others! > > My setup : I run Xen PV guests. LOTS of them. What I do to ease their > management is use unpartitioned LVM logical volumes from the host as > block devices, one for "/" and the other for swap. koan already supports specifying multiple virtual disks, which can use LVM. So you're ok there. (--virt-path) > I don't let any > tools create the guest config files, I create them myself "by hand" > once, then deploy them through an rpm package to all the Xen hosts I > have available. This way I have a central location to check for all > guest configuration and MAC addresses. > Cobbler is designed for central configuration, so it can be your MAC database. > The initial guest install is where I'd like to use cobbler and koan in > order to save time. I currently extract and modify a pre-existing RHEL5 > install, but that's another story ;-) > > Problem : Koan wants to create the Xen guest configuration files for me. > I'd like it to be able to use an existing file and have koan extract the > settings it needs from there. > I'm not interested in adding this because then we do not leverage the full power of libvirt ... we can configure /different/ types of virtualization without having to hardware in Xen specific features. > Here is an example Xen guest config file I use : > > --- > # Xen Configuration File > name = "cobbler" > memory = "512" > maxmem = "512" > disk = [ 'phy:/dev/data/cobbler,xvda,w', > 'phy:/dev/data/swapcobbler,xvdb,w' ] > vif = [ 'mac=00:16:3e:5f:76:6e, bridge=xenbr0', > 'mac=00:16:3e:5f:76:6f, bridge=xenbr1' ] > uuid = '5f766eef-8cf3-c287-1635-115e103561d0' > > bootloader = '/usr/bin/pygrub' > vcpus = 2 > on_reboot = 'restart' > on_crash = 'restart' > --- > All of the above can be set in koan. Well, all but the multiple bridges. Currently you can pick one. For 0.6.3, you'll be able to specify you'd like more than one, though I need to figure out how I want to handle this from the CLI perspective to make it sufficiently easy to understand. > I generate the uuid myself (md5sum of the full host name), then use the > 6 first characters in the eth0 mac address, increment it for the eth1 > mac address. > > Ideally, I'd like to be able to tell koan to use this existing > "cobbler" file and use xenbr1 for everything it needs to do (since > dhcpd and cobbler are only listening on that "private" network), then > wait, and get a working "cobbler" system up and running :-) > I'm not sure what you mean with regards to "existing cobbler file". Cobbler already has mechanisms for specifying attributes related to virtualization. > Is this something that could be achieved? Am I the only one to have > multiple block devices, multiple network interfaces, and like having > fixed MAC addresses for my Xen guests? > Cobbler already supports multiple block devices and assigning specific MACs. The only thing it currently lacks from what you list above, that I can see, is the ability to specify multiple NICs, which I plan to address in the next release. > Currently, koan might be useful to me, but once I get the guest > running, I'd still need to move stuff around block devices and change > bits and pieces of configuration to get things into place as I want > them. > Yep, thanks for sharing... this now confirms there are at least two people who want control over two NICs, so it will be surely done now :) My main concern still is in expressing this configuration in a simple way without overly complicating the --virt-bridge parameter. It shouldn't be too difficult. Anyhow, look at the other features in the koan manpage and cobbler documentation to see how to do LVM, multiple disks, and so on. > Matthias > > From msinhore at gmail.com Wed Sep 26 17:56:01 2007 From: msinhore at gmail.com (Marco Sinhoreli) Date: Wed, 26 Sep 2007 14:56:01 -0300 Subject: [et-mgmt-tools] python libvirt methods to virDomainBlockStats and virDomainInterfaceStats Message-ID: <20fe3cf60709261056v44046d39hadaa152a3ecf0fa5@mail.gmail.com> Hello all, I'm using libvirt from cvs, and in libvirt.h there are virDomainBlockStats, and virDomainInterfaceStats calls. I think it is compiled into libvirtmod, then I'm trying to write two methods, blockStats, and interfacesStats in the class virDomain at the libvirt.py library file. I'm not a specialist in python programming and it's not working. Bellow folow the code I have been wrote: def blockStats(self): """Block device stats for virDomainBlockStats """ ret = libvirtmod.virDomainBlockStats(self._o) if ret is None: raise libvirtError ('virDomainBlockStats() failed', dom=self) return ret def interfaceStats(self): """Network interface stats for virDomainInterfaceStats.""" ret = libvirtmod.virDomainInterfaceStats(self._o) if ret is None: raise libvirtError ('virDomainInterfaceStats() failed', dom=self) return ret Any help will be welcome. Best regards -- Marco Sinhoreli From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Wed Sep 26 17:58:16 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Wed, 26 Sep 2007 19:58:16 +0200 Subject: [et-mgmt-tools] Cobbler WebUI log in problem Message-ID: <20070926195816.16b577fc@python3.es.egwn.lan> Hi again, Still using the 0.6.2 cobbler pre-release from today, I restarted cobblerd while logged in at the WebUI, but then I couldn't get it to work anymore : I would log in, but then clicking on any of the menu items would show the login screen again and again. Attached are the logs from when that happens. The "fix" was to restart my browser. The two last lines are the successful login after having done that. Other minor annoyance : When clicking on "SYNC" in the WebUI, the usual output from "cobbler sync" is displayed in the terminal where "service cobblerd start" was run. That output should probably be caught by the WebUI and displayed to the user instead (if it's possible). Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.19 0.29 0.37 -------------- next part -------------- An embedded and charset-unspecified text was scrubbed... Name: cobbler-logs.txt URL: From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Wed Sep 26 18:11:10 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Wed, 26 Sep 2007 20:11:10 +0200 Subject: [et-mgmt-tools] Cobbler and Koan for _my_ needs In-Reply-To: <46FA99C5.6010806@redhat.com> References: <20070926185843.5f040191@python3.es.egwn.lan> <46FA99C5.6010806@redhat.com> Message-ID: <20070926201110.6a559578@python3.es.egwn.lan> Michael DeHaan wrote : > koan already supports specifying multiple virtual disks, which can use LVM. > So you're ok there. > > (--virt-path) Neat. I completely missed that, it seems. > Cobbler is designed for central configuration, so it can be your MAC > database. Fair enough :-) I'll need to figure out the best way to have it "store" the two MAC addresses for my Xen guests. I guess that's not possible yet given your other comments, so when you implement the multiple bridges support, please don't forget me! ;-) [...] > Yep, thanks for sharing... this now confirms there are at least two > people who want > control over two NICs, so it will be surely done now :) My main concern > still is in expressing > this configuration in a simple way without overly complicating the > --virt-bridge parameter. > It shouldn't be too difficult. > > Anyhow, look at the other features in the koan manpage and cobbler > documentation to see how > to do LVM, multiple disks, and so on. Thanks for your answer! I'll poke cobbler and koan some more, especially koan since you told me that the bug preventing it from using xenbr1 was fixed in 0.6.2, so I should be able to get further with it now, even if it can't (yet) set up the two vifs I need automatically. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.48 0.48 0.42 From wright at imageworks.com Wed Sep 26 18:11:52 2007 From: wright at imageworks.com (Peter Wright) Date: Wed, 26 Sep 2007 11:11:52 -0700 Subject: [et-mgmt-tools] Cobbler WebUI log in problem In-Reply-To: <20070926195816.16b577fc@python3.es.egwn.lan> References: <20070926195816.16b577fc@python3.es.egwn.lan> Message-ID: <46FAA0E8.9000806@imageworks.com> Matthias Saou wrote: > > Hi again, > > Still using the 0.6.2 cobbler pre-release from today, I restarted > cobblerd while logged in at the WebUI, but then I couldn't get it to > work anymore : I would log in, but then clicking on any of the menu > items would show the login screen again and again. > > Attached are the logs from when that happens. The "fix" was to restart > my browser. The two last lines are the successful login after having > done that. > yea i had that happen as well. i just removed the cookies issued from my node running the webUI. -p -- Peter Wright Systems Administrator Sony Pictures Imageworks wright at imageworks.com www.imageworks.com From corey.osman at turner.com Wed Sep 26 18:13:15 2007 From: corey.osman at turner.com (Corey Osman) Date: Wed, 26 Sep 2007 14:13:15 -0400 Subject: [et-mgmt-tools] dhcp configuration options Message-ID: Hi, I have a large datacenter about 2000 systems or more so my dhcpd.conf is very large (1800 lines). I played around with the cobbler dhcp management but I don't like they way it inserts the data. here is a small section of my dhcpd.conf file. There are lots of groups in the file. group { option routers 192.168.5.1; option subnet-mask 255.255.255.0; host linuxhost1 { hardware ethernet 00:02:0D:60:12:0C; fixed- address 192.168.5.48; } host linuxhost12 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.49; } host linuxhost1y1 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.50; } host paylinuxhost1 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.51; } host linuxhost3 { hardware ethernet 00:02:0D:60:12:0C; fixed- address 192.168.5.56; } host linuxhostcvs1 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.69; } host linuxhost-db1 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.139; } host linuxhost-stg2 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.130; } host linuxhost-ref1 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.131; } host linuxhost-stg1 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.129; } host linuxhost-app1 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.132; } host linuxhostw1 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.133; } host linuxhost { hardware ethernet 00:02:0D:60:12:0C; fixed- address 192.168.5.134; } host linuxhost-app1 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.136; } host linuxhost-app2 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.137; } host linuxhost-db2 { hardware ethernet 00:02:0D:60:12:0C; fixed-address 192.168.5.138; } } When cobbler adds an entry it does the following: host label1 { hardware ethernet 00:50:56:ae:44:c5; fixed-address 192.167.5.202; next-server 192.168.0.1; } Now, this is nice but when you have 2000 entries it gets quite messy. Will the commands "cobbler sync" and "cobbler list" be able to handle the large amount of entries? Does anyone know how I can get cobbler to: A. change label1 to the hostname B. Get cobbler to remove the new lines C. Determine which group to add the entry too based on IP address So that it just adds a line like the following to the correct group host linuxhost-db2 { hardware ethernet 00:02:0D:60:12:0C; fixed- address 192.168.5.138; } From mdehaan at redhat.com Wed Sep 26 18:15:44 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 26 Sep 2007 14:15:44 -0400 Subject: [et-mgmt-tools] dhcp configuration options In-Reply-To: References: Message-ID: <46FAA1D0.6060804@redhat.com> Corey Osman wrote: > Hi, > > I have a large datacenter about 2000 systems or more so my dhcpd.conf > is very large (1800 lines). I played around with the cobbler dhcp > management but I don't like they way it inserts the data. > Reply below... > > here is a small section of my dhcpd.conf file. There are lots of > groups in the file. > > group { > option routers 192.168.5.1; > option subnet-mask 255.255.255.0; > host linuxhost1 { hardware ethernet 00:02:0D:60:12:0C; fixed-address > 192.168.5.48; } > host linuxhost12 { hardware ethernet 00:02:0D:60:12:0C; fixed-address > 192.168.5.49; } > host linuxhost1y1 { hardware ethernet 00:02:0D:60:12:0C; fixed-address > 192.168.5.50; } > host paylinuxhost1 { hardware ethernet 00:02:0D:60:12:0C; > fixed-address 192.168.5.51; } > host linuxhost3 { hardware ethernet 00:02:0D:60:12:0C; fixed-address > 192.168.5.56; } > host linuxhostcvs1 { hardware ethernet 00:02:0D:60:12:0C; > fixed-address 192.168.5.69; } > host linuxhost-db1 { hardware ethernet 00:02:0D:60:12:0C; > fixed-address 192.168.5.139; } > host linuxhost-stg2 { hardware ethernet 00:02:0D:60:12:0C; > fixed-address 192.168.5.130; } > host linuxhost-ref1 { hardware ethernet 00:02:0D:60:12:0C; > fixed-address 192.168.5.131; } > host linuxhost-stg1 { hardware ethernet 00:02:0D:60:12:0C; > fixed-address 192.168.5.129; } > host linuxhost-app1 { hardware ethernet 00:02:0D:60:12:0C; > fixed-address 192.168.5.132; } > host linuxhostw1 { hardware ethernet 00:02:0D:60:12:0C; fixed-address > 192.168.5.133; } > host linuxhost { hardware ethernet 00:02:0D:60:12:0C; fixed-address > 192.168.5.134; } > host linuxhost-app1 { hardware ethernet 00:02:0D:60:12:0C; > fixed-address 192.168.5.136; } > host linuxhost-app2 { hardware ethernet 00:02:0D:60:12:0C; > fixed-address 192.168.5.137; } > host linuxhost-db2 { hardware ethernet 00:02:0D:60:12:0C; > fixed-address 192.168.5.138; } > } > > When cobbler adds an entry it does the following: > > host label1 { > hardware ethernet 00:50:56:ae:44:c5; > fixed-address 192.167.5.202; > next-server 192.168.0.1; > } > You're running an older cobbler version, it seems. In newer cobbler, if you have the "hostname" field set on the object, label (which I've since changed to just say "generic") will be replaced by the hostname in use. As it appears you do have hostname info (at least in your manual setup), hopefully this works for you. The duplicate next-server entry is also removed in more recent versions. I can't see anything against having a more consolidated file .... seems like a sane thing to do to me. > Now, this is nice but when you have 2000 entries it gets quite messy. > Will the commands "cobbler sync" and "cobbler list" be able to handle > the large amount of entries? > Yes. They should just take longer :) > > Does anyone know how I can get cobbler to: > A. change label1 to the hostname See above. > B. Get cobbler to remove the new lines That's going to require a code change, but it can be done. > C. Determine which group to add the entry too based on IP address Essentially: https://hosted.fedoraproject.org/projects/cobbler/wiki/UsingCobblerDhcpManagementWithMultipleSubnets > > So that it just adds a line like the following to the correct group > > host linuxhost-db2 { hardware ethernet 00:02:0D:60:12:0C; > fixed-address 192.168.5.138; } > > _______________________________________________ > 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 Wed Sep 26 18:30:12 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 26 Sep 2007 14:30:12 -0400 Subject: [et-mgmt-tools] Reminder about where you're getting your Cobbler Message-ID: <46FAA534.4080208@redhat.com> Hi, So I've seen a lot of questions today on IRC and email from things that look like older versions of Cobbler. Many questions raised have been things that are new features (or changes) that are present in newer versions. So, upgrading would solve a lot :) The best place to get the latest Cobbler is always going to be Fedora (via yum) or EPEL testing (which also works for yum). Use the following URL as the mirrors may not in sync: http://download.fedora.redhat.com/pub/epel/testing/ If you are using Centos Extras, you have some fairly old bits and should also look into using the above repos, which are compatible. EPEL testing doesn't mean the code is "testing" grade, it just means I can update it as needed as opposed to whenever the EPEL maintainers decide to move packages over from testing (this used to be about quarterly, not sure what it is now). Thanks! --Michael From msinhore at gmail.com Wed Sep 26 18:48:04 2007 From: msinhore at gmail.com (Marco Sinhoreli) Date: Wed, 26 Sep 2007 15:48:04 -0300 Subject: [et-mgmt-tools] Re: python libvirt methods to virDomainBlockStats and virDomainInterfaceStats In-Reply-To: <20fe3cf60709261056v44046d39hadaa152a3ecf0fa5@mail.gmail.com> References: <20fe3cf60709261056v44046d39hadaa152a3ecf0fa5@mail.gmail.com> Message-ID: <20fe3cf60709261148j4193f663t165e39ddb9be3723@mail.gmail.com> Complementing the last mail: Importing the libvirtmod the virDomainInterfaceStats, and virDomainBlockStats aren't listed (using help(libvirtmod)) Regards On 9/26/07, Marco Sinhoreli wrote: > Hello all, > > I'm using libvirt from cvs, and in libvirt.h there are > virDomainBlockStats, and virDomainInterfaceStats calls. I think it is > compiled into libvirtmod, then I'm trying to write two methods, > blockStats, and interfacesStats in the class virDomain at the > libvirt.py library file. I'm not a specialist in python programming > and it's not working. > > Bellow folow the code I have been wrote: > > def blockStats(self): > """Block device stats for virDomainBlockStats """ > ret = libvirtmod.virDomainBlockStats(self._o) > if ret is None: raise libvirtError ('virDomainBlockStats() > failed', dom=self) > return ret > > def interfaceStats(self): > """Network interface stats for virDomainInterfaceStats.""" > ret = libvirtmod.virDomainInterfaceStats(self._o) > if ret is None: raise libvirtError ('virDomainInterfaceStats() > failed', dom=self) > return ret > > Any help will be welcome. > > Best regards > > -- > Marco Sinhoreli > -- Marco Sinhoreli From marcmo at foxriver.com Wed Sep 26 18:50:26 2007 From: marcmo at foxriver.com (Marc Mondragon) Date: Wed, 26 Sep 2007 13:50:26 -0500 Subject: [et-mgmt-tools] WebUI issue Message-ID: Hello all, I've been playing around with the new WebUI and everything seems to be working fine. However, when I click on the Add->System link I get a page of highlighted code with this at the end: global ParseError = , self = , msg undefined ParseError: Invalid end directive Line 107, column 9 in file /usr/share/cobbler/webui_templates/system_edit.tmpl Line|Cheetah Code ----|------------------------------------------------------------- 104 | /> 105 | #if not $system 106 | random 107 | #end ^ 108 |

Example: AA:BB:CC:DD:EE:FF

109 | 110 | col = None extMsg = '' lineno = None msg = 'Invalid end directive' report = > stream = I did update this afternoon from git and I haven't added any systems recently so I cannot recall if this was happening before the update. Marc Mondragon Fox River Financial Resources/Ritchie Capital Investments, Ltd. 2100 Enterprise Avenue Geneva, IL 60134 marcmo at foxriver.com -------------- next part -------------- An HTML attachment was scrubbed... URL: From mdehaan at redhat.com Wed Sep 26 18:51:41 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 26 Sep 2007 14:51:41 -0400 Subject: [et-mgmt-tools] WebUI issue In-Reply-To: References: Message-ID: <46FAAA3D.3030704@redhat.com> Marc Mondragon wrote: > > Hello all, > > I?ve been playing around with the new WebUI and everything seems to be > working fine. However, when I click on the Add->System link I get a > page of highlighted code with this at the end: > > //global// **ParseError** = , > **self** = , msg //undefined// > > **ParseError**: Invalid end directive Line 107, column 9 in file > /usr/share/cobbler/webui_templates/system_edit.tmpl Line|Cheetah Code > ----|------------------------------------------------------------- 104 > | /> 105 | #if not $system 106 | random 107 | #end ^ > 108 |

Example: AA:BB:CC:DD:EE:FF

109 | > 110 | > col = None > extMsg = '' > lineno = None > msg = 'Invalid end directive' > report = instance>> > stream = > > I did update this afternoon from git and I haven?t added any systems > recently so I cannot recall if this was happening before the update. > > Marc Mondragon > > Fox River Financial Resources/Ritchie Capital Investments, Ltd. > 2100 Enterprise Avenue > Geneva, IL 60134 > marcmo at foxriver.com > > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools This is fixed in the patch I linked earlier today. Namely, there was a stray "#" in one of the templates. If you want to pull from git and retest this should be happy now. Appreciate the help in looking things over. --Michael From msinhore at gmail.com Wed Sep 26 19:05:21 2007 From: msinhore at gmail.com (Marco Sinhoreli) Date: Wed, 26 Sep 2007 16:05:21 -0300 Subject: [et-mgmt-tools] Re: python libvirt methods to virDomainBlockStats and virDomainInterfaceStats In-Reply-To: <20fe3cf60709261148j4193f663t165e39ddb9be3723@mail.gmail.com> References: <20fe3cf60709261056v44046d39hadaa152a3ecf0fa5@mail.gmail.com> <20fe3cf60709261148j4193f663t165e39ddb9be3723@mail.gmail.com> Message-ID: <20fe3cf60709261205r577269dh89db9a7c5ce05952@mail.gmail.com> Investigating the source, I find the libvirtclass.txt file in libvirt-source/python directory with these contents: [...] Functions failed: virConnSetErrorFunc virDomainGetSchedulerParameters virConnectGetVersion virSetErrorFunc virDomainSetSchedulerParameters virConnCopyLastError virDomainBlockStats virCopyLastError virDomainGetVcpus virDefaultErrorFunc virResetError virDomainInterfaceStats On 9/26/07, Marco Sinhoreli wrote: > Complementing the last mail: > Importing the libvirtmod the virDomainInterfaceStats, and > virDomainBlockStats aren't listed (using help(libvirtmod)) > > Regards > > On 9/26/07, Marco Sinhoreli wrote: > > Hello all, > > > > I'm using libvirt from cvs, and in libvirt.h there are > > virDomainBlockStats, and virDomainInterfaceStats calls. I think it is > > compiled into libvirtmod, then I'm trying to write two methods, > > blockStats, and interfacesStats in the class virDomain at the > > libvirt.py library file. I'm not a specialist in python programming > > and it's not working. > > > > Bellow folow the code I have been wrote: > > > > def blockStats(self): > > """Block device stats for virDomainBlockStats """ > > ret = libvirtmod.virDomainBlockStats(self._o) > > if ret is None: raise libvirtError ('virDomainBlockStats() > > failed', dom=self) > > return ret > > > > def interfaceStats(self): > > """Network interface stats for virDomainInterfaceStats.""" > > ret = libvirtmod.virDomainInterfaceStats(self._o) > > if ret is None: raise libvirtError ('virDomainInterfaceStats() > > failed', dom=self) > > return ret > > > > Any help will be welcome. > > > > Best regards > > > > -- > > Marco Sinhoreli > > > > > -- > Marco Sinhoreli > -- Marco Sinhoreli From berrange at redhat.com Thu Sep 27 00:37:34 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 27 Sep 2007 01:37:34 +0100 Subject: [et-mgmt-tools] Adding / removing mice, tablets & displays in virt-manager Message-ID: <20070927003734.GK29729@redhat.com> Code I have just pushed to the virt-manager mercurial repository has extended the 'Add hardware' dialog to allow you to add mice, graphics tablets and displays to a VM. Adding mice isn't particularly useful, but adding a graphics tablet enables you to get a mouse which tracks 1-for-1 between the guest and the host[1]. Adding a display allows you to take an existing VM which was installed in text mode, and enable VNC support in it. With VNC support the graphic console will appear within virt-manager, and with suitable config can also be accessed from a remote host. Regards, Dan. [1] Windows will configure it automatically. Linux requires use of 'evdev' driver in xorg.conf -- |=- 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 kunkichi at gmail.com Thu Sep 27 00:50:24 2007 From: kunkichi at gmail.com (Shimizu Kuniaki) Date: Thu, 27 Sep 2007 09:50:24 +0900 Subject: [et-mgmt-tools] Abount Japanese Translation for Cobbler Documents Message-ID: <4dd11c330709261750q7925556bo4d437bdee5916cef@mail.gmail.com> Hi, As using cobbler, I think Cobbler is a very nice tool, but documents are available only in English. I would like to translate some cobbler documents into Japanese and put on my website. Is that OK? Kuniaki Shimizu kunkichi at gmail.com From mdehaan at redhat.com Thu Sep 27 02:30:54 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 26 Sep 2007 22:30:54 -0400 Subject: [et-mgmt-tools] Abount Japanese Translation for Cobbler Documents In-Reply-To: <4dd11c330709261750q7925556bo4d437bdee5916cef@mail.gmail.com> References: <4dd11c330709261750q7925556bo4d437bdee5916cef@mail.gmail.com> Message-ID: <46FB15DE.7090706@redhat.com> Shimizu Kuniaki wrote: > Hi, > > As using cobbler, I think Cobbler is a very nice tool, but documents > are available only in English. > I would like to translate some cobbler documents into Japanese and put > on my website. > Is that OK? > > Kuniaki Shimizu > kunkichi at gmail.com > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools > That's great and I'll gladly link to the translations from the main page. Thanks! --Michael From pcbailey37 at hotmail.com Thu Sep 27 02:34:15 2007 From: pcbailey37 at hotmail.com (Patrice Bailey) Date: Wed, 26 Sep 2007 21:34:15 -0500 Subject: [et-mgmt-tools] please remove me from list In-Reply-To: <021b01c7f488$b9342380$3b96b5c6@igneous.opin.com> References: <34a912120708271122l4c491ddndc6f1ef18de891b2@mail.gmail.com> <34a912120709010655v3a72788fsce1eb77c0689dd16@mail.gmail.com> <014101c7ee27$8bef00a0$6737a8c0@igneous.opin.com> <021b01c7f488$b9342380$3b96b5c6@igneous.opin.com> Message-ID: Can you please remove me from your email list. Thank youPatrice C. Bailey -------------- next part -------------- An HTML attachment was scrubbed... URL: From mdehaan at redhat.com Thu Sep 27 02:37:30 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Wed, 26 Sep 2007 22:37:30 -0400 Subject: [et-mgmt-tools] please remove me from list In-Reply-To: References: <34a912120708271122l4c491ddndc6f1ef18de891b2@mail.gmail.com> <34a912120709010655v3a72788fsce1eb77c0689dd16@mail.gmail.com> <014101c7ee27$8bef00a0$6737a8c0@igneous.opin.com> <021b01c7f488$b9342380$3b96b5c6@igneous.opin.com> Message-ID: <46FB176A.6020306@redhat.com> Patrice Bailey wrote: > Can you please remove me from your email list. Thank you > > /Patrice C. Bailey/ > ------------------------------------------------------------------------ > > _______________________________________________ > et-mgmt-tools mailing list > et-mgmt-tools at redhat.com > https://www.redhat.com/mailman/listinfo/et-mgmt-tools This is a self-service list, so you can use the link at the bottom of every email to change your subscription preferences. https://www.redhat.com/mailman/listinfo/et-mgmt-tools --Michael From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Thu Sep 27 08:56:16 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Thu, 27 Sep 2007 10:56:16 +0200 Subject: [et-mgmt-tools] Cobbler and Koan for _my_ needs In-Reply-To: <46FA99C5.6010806@redhat.com> References: <20070926185843.5f040191@python3.es.egwn.lan> <46FA99C5.6010806@redhat.com> Message-ID: <20070927105616.3d111c9f@python3.es.egwn.lan> Michael DeHaan wrote : > Cobbler is designed for central configuration, so it can be your MAC > database. Can it really? For Xen guests? I've played some more with Koan and Xen guests, but apparently even telling Koan to install a system which has a specific MAC address in its Cobbler system details creates the Xen guest file with a random MAC address. My DHCP server has only static entries, which means that if Koan used the MAC address I've put in the Cobbler system it would work, but the random MAC address won't ever be able to get an IP address (and this corporate network _must_ stay this way). Using Koan's "--display", it does seem like the MAC address isn't obtained from Cobbler. Am I maybe simply doing something wrong? On a related note, I do see "virt_path" and "virt_type" in the Cobbler systems, but I don't see "virt_ram", which would be very useful to be able to set on a per-system basis in my case. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.33 0.41 0.42 From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Thu Sep 27 11:02:59 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Thu, 27 Sep 2007 13:02:59 +0200 Subject: [et-mgmt-tools] WebUI issue In-Reply-To: <46FAAA3D.3030704@redhat.com> References: <46FAAA3D.3030704@redhat.com> Message-ID: <20070927130259.7a4e58fc@python3.es.egwn.lan> Michael DeHaan wrote : > This is fixed in the patch I linked earlier today. > Namely, there was a stray "#" in one of the templates. > If you want to pull from git and retest this should be happy now. > > Appreciate the help in looking things over. The code I got from git when you wrote that this problem should have been fixed is still wrong. Apart from the s/random#/random/, you also moved the #end back to its own line, but "#end" isn't correct, it needs to be "#end if" for it to work apparently (I've just tested this). Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.29 0.37 0.37 From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Thu Sep 27 11:14:17 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Thu, 27 Sep 2007 13:14:17 +0200 Subject: [et-mgmt-tools] WebUI issue In-Reply-To: <20070927130259.7a4e58fc@python3.es.egwn.lan> References: <46FAAA3D.3030704@redhat.com> <20070927130259.7a4e58fc@python3.es.egwn.lan> Message-ID: <20070927131417.51880424@python3.es.egwn.lan> Matthias Saou wrote : > Michael DeHaan wrote : > > > This is fixed in the patch I linked earlier today. > > Namely, there was a stray "#" in one of the templates. > > If you want to pull from git and retest this should be happy now. > > > > Appreciate the help in looking things over. > > The code I got from git when you wrote that this problem should have > been fixed is still wrong. Apart from the s/random#/random/, you also > moved the #end back to its own line, but "#end" isn't correct, it needs > to be "#end if" for it to work apparently (I've just tested this). Replying to myself, but... With this now fixed, I get the ADD->SYSTEM page to display, but when I click on the "random" link for the MAC address, the field gets filled up with HTML code. Then when I enter a MAC address myself, and try to save, I get a page with this at the bottom : fault- \\n args =\\n()\\n faultcode =\\n1\\n faultstring =\\n"cobbler.cexceptions.cx-\\\'queried object has no field- mac\\\'"\\n\\n\\n\''> Also, when I tried earlier to edit an existing SYSTEM I had added with the command-line, to remove the IP address I had set, it wouldn't go away. Same for the hostname for instance. Right now, these errors seem to have broken my setup pretty badly, since the system I added above got this in the "systems" file : mac_address: "AA:BB:CC:DD:EE:FF\n References: <46FBA02A.4040504@inria.fr> Message-ID: <46FBAA69.90801@redhat.com> Guillaume Rousse wrote: > virtmanager split content of /etc/sysconfig/keyboard on '"' for > determining keyboard. On mandriva system, there isn't any quote in this > file: see http://qa.mandriva.com/show_bug.cgi?id=34190 > > Here is a patch making parsing of this file a bit more robust. Patch forwarded to Fedora/Linux Management Tools list. 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: virt-manager-0.5.0.keyboard.patch Type: text/x-patch Size: 782 bytes Desc: not available URL: -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 3237 bytes Desc: S/MIME Cryptographic Signature URL: From mdehaan at redhat.com Thu Sep 27 14:26:04 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 27 Sep 2007 10:26:04 -0400 Subject: [et-mgmt-tools] Cobbler and Koan for _my_ needs In-Reply-To: <20070927105616.3d111c9f@python3.es.egwn.lan> References: <20070926185843.5f040191@python3.es.egwn.lan> <46FA99C5.6010806@redhat.com> <20070927105616.3d111c9f@python3.es.egwn.lan> Message-ID: <46FBBD7C.7010504@redhat.com> Matthias Saou wrote: > Michael DeHaan wrote : > > >> Cobbler is designed for central configuration, so it can be your MAC >> database. >> > > Can it really? For Xen guests? > Yes. koan --virt --system=AA:BB:CC:DD:EE:FF --server=cobbler.example.org > I've played some more with Koan and Xen guests, but apparently even > telling Koan to install a system which has a specific MAC address in > its Cobbler system details creates the Xen guest file with a random MAC > address. What cobbler and koan version is this? > My DHCP server has only static entries, which means that if > Koan used the MAC address I've put in the Cobbler system it would work, > but the random MAC address won't ever be able to get an IP address (and > this corporate network _must_ stay this way). > > Using Koan's "--display", it does seem like the MAC address isn't > obtained from Cobbler. > > Am I maybe simply doing something wrong? > > On a related note, I do see "virt_path" and "virt_type" in the Cobbler > systems, but I don't see "virt_ram", which would be very useful to be > able to set on a per-system basis in my case. > Conceptually, profiles are there to describe what the purpose of the system is and it's requirements. If a profile states that a configuration needs "X" amount of RAM, it is important to honor that. The overrides on storage are for cases such as a development profile being created centrally, and the local user wanting to install it in a different location (for instance, say they had a large partition on the domU they wanted to use). Overriding the amount of virtual RAM on the koan command line is possible, but we do not want folks thinking of RAM requirements for individual systems because that defeats the point of the profile abstraction. > Matthias > > From mdehaan at redhat.com Thu Sep 27 14:27:24 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 27 Sep 2007 10:27:24 -0400 Subject: [et-mgmt-tools] WebUI issue In-Reply-To: <20070927130259.7a4e58fc@python3.es.egwn.lan> References: <46FAAA3D.3030704@redhat.com> <20070927130259.7a4e58fc@python3.es.egwn.lan> Message-ID: <46FBBDCC.1000606@redhat.com> Matthias Saou wrote: > Michael DeHaan wrote : > > >> This is fixed in the patch I linked earlier today. >> Namely, there was a stray "#" in one of the templates. >> If you want to pull from git and retest this should be happy now. >> >> Appreciate the help in looking things over. >> > > The code I got from git when you wrote that this problem should have > been fixed is still wrong. Apart from the s/random#/random/, you also > moved the #end back to its own line, but "#end" isn't correct, it needs > to be "#end if" for it to work apparently (I've just tested this). > > Matthias > > Thanks. I'm leaving most of the WebUI changes up to Al at this point, but wanted to get the quick fix out there for those that wanted to play with it. I'll correct it. From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Thu Sep 27 15:03:06 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Thu, 27 Sep 2007 17:03:06 +0200 Subject: [et-mgmt-tools] Cobbler and Koan for _my_ needs In-Reply-To: <46FBBD7C.7010504@redhat.com> References: <20070926185843.5f040191@python3.es.egwn.lan> <46FA99C5.6010806@redhat.com> <20070927105616.3d111c9f@python3.es.egwn.lan> <46FBBD7C.7010504@redhat.com> Message-ID: <20070927170306.1d5f20f7@python3.es.egwn.lan> Michael DeHaan wrote : > >> Cobbler is designed for central configuration, so it can be your MAC > >> database. > > > > Can it really? For Xen guests? > > Yes. > > koan --virt --system=AA:BB:CC:DD:EE:FF --server=cobbler.example.org Hmmmm, indeed, this works, but this makes things uglier for me. Renaming my system from 'test' to '00:16:3e:01:ec:2e' in Cobbler and using --system=00:16:3e:01:ec:2e instead of --system=test makes the virtual guest get the intended the MAC address, but : - I now have a system called '00:16:3e:01:ec:2e' in the WebUI instead of the much more human friendly 'test'. - I now have a Xen guest called '00_16_3e_01_ec_2e' instead of the much more human friendly 'test'. When I run 'xm list' and see a bunch of host names, it's much quicker for me to stop/restart/connect-to any of the guests than with having a list of 00_16_3e_xx_xx_xx entries :-/ Same for the WebUI... So it seems to me that there is too much logic on the Koan side, which makes it impossible to use a friendly name and have it use the MAC address configured in the Cobbler. > > I've played some more with Koan and Xen guests, but apparently even > > telling Koan to install a system which has a specific MAC address in > > its Cobbler system details creates the Xen guest file with a random MAC > > address. > > What cobbler and koan version is this? The 0.6.2-1 you posted yesterday. But see above : Invoking the system by "friendly name" gets a random MAC address, even if the system has a MAC address configured in Cobbler, but renaming the system to its MAC address "fixes" this. > > My DHCP server has only static entries, which means that if > > Koan used the MAC address I've put in the Cobbler system it would work, > > but the random MAC address won't ever be able to get an IP address (and > > this corporate network _must_ stay this way). > > > > Using Koan's "--display", it does seem like the MAC address isn't > > obtained from Cobbler. > > > > Am I maybe simply doing something wrong? > > > > On a related note, I do see "virt_path" and "virt_type" in the Cobbler > > systems, but I don't see "virt_ram", which would be very useful to be > > able to set on a per-system basis in my case. > > > > Conceptually, profiles are there to describe what the purpose of the > system is and it's requirements. > If a profile states that a configuration needs "X" amount of RAM, it is > important to honor that. > > The overrides on storage are for cases such as a development profile > being created centrally, and the local user > wanting to install it in a different location (for instance, say they > had a large partition on the domU they wanted > to use). > > Overriding the amount of virtual RAM on the koan command line is > possible, but we do not want folks > thinking of RAM requirements for individual systems because that defeats > the point of the profile > abstraction. I don't want to override anything on the koan command line. I'm more than fine with having everything in Cobbler! But in my case, I'd need my profile to be : - virt_path: /dev/data/$name,/dev/data/swap$name (I'm not sure if this can be done, I can live with overriding it in all system configurations, which is what I've done for now) - virt_ram: 512 Then the default system associated to be : - virt_path: <> - virt_ram: <> ...where I only change the name, hostname and MAC address. Here I'd like to be able to change virt_ram on a per-system basis, but it doesn't seem possible currently (or maybe I just need to "manually" add it to the system file?). In my scenario, we're not talking about thousands of servers, more like a few dozens in each location, with locations being very remote and having each their own cobbler instance, so I'm thinking about creating a system entry in Cobbler for each and every host and guest I have. Obviously, they'll all inherit most of their configuration from a given profile, but I'll need to override some settings here and there. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.54 0.46 0.37 From mdehaan at redhat.com Thu Sep 27 15:20:41 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 27 Sep 2007 11:20:41 -0400 Subject: [et-mgmt-tools] Cobbler and Koan for _my_ needs In-Reply-To: <20070927170306.1d5f20f7@python3.es.egwn.lan> References: <20070926185843.5f040191@python3.es.egwn.lan> <46FA99C5.6010806@redhat.com> <20070927105616.3d111c9f@python3.es.egwn.lan> <46FBBD7C.7010504@redhat.com> <20070927170306.1d5f20f7@python3.es.egwn.lan> Message-ID: <46FBCA49.6070500@redhat.com> Matthias Saou wrote: > Michael DeHaan wrote : > > >>>> Cobbler is designed for central configuration, so it can be your MAC >>>> database. >>>> >>> Can it really? For Xen guests? >>> >> Yes. >> >> koan --virt --system=AA:BB:CC:DD:EE:FF --server=cobbler.example.org >> > > Hmmmm, indeed, this works, but this makes things uglier for me. > Renaming my system from 'test' to '00:16:3e:01:ec:2e' in Cobbler and > using --system=00:16:3e:01:ec:2e instead of --system=test makes the > virtual guest get the intended the MAC address, but : > - I now have a system called '00:16:3e:01:ec:2e' in the WebUI instead > of the much more human friendly 'test'. > - I now have a Xen guest called '00_16_3e_01_ec_2e' instead of the much > more human friendly 'test'. > > Hmm, this looks like leftover bits from when the name had to be a MAC. Since the MAC is a seperate field now, I'll see why it's not being used if it's provided. However, even so, you can still add the optional parameter --virt-name to name your guest. koan --virt --system=AA:BB:CC:DD:EE:FF --server=cobbler.example.org --virt-name=blinky > When I run 'xm list' and see a bunch of host names, it's much quicker > for me to stop/restart/connect-to any of the guests than with having a > list of 00_16_3e_xx_xx_xx entries :-/ Same for the WebUI... > See above... > So it seems to me that there is too much logic on the Koan side, which > makes it impossible to use a friendly name and have it use the MAC > address configured in the Cobbler. > Nope... even if the "find out what the MAC address is set to" is currently fouled up for virtual machines when the name is not a MAC, you still have a way to specify a name override. > I don't want to override anything on the koan command line. I'm more > than fine with having everything in Cobbler! > > But in my case, I'd need my profile to be : > - virt_path: /dev/data/$name,/dev/data/swap$name > (I'm not sure if this can be done, I can live with overriding it in > all system configurations, which is what I've done for now) > - virt_ram: 512 > This would put them in two directories, which should be sufficient: --virt-path=/dev/data,/dev/data/swap/ > Then the default system associated to be : > - virt_path: <> > - virt_ram: <> > ...where I only change the name, hostname and MAC address. > > Here I'd like to be able to change virt_ram on a per-system basis, but > it doesn't seem possible currently (or maybe I just need to "manually" > add it to the system file?). > It's not like we couldn't add it, it just seems to go against the design of having profiles that define the requirements of what they are running. This is the same reason you can't request more virtual disk for a specific system -- the idea is that you would use inherited profiles to do this, and then just map the system to the profile that it was going to be assigned to. > In my scenario, we're not talking about thousands of servers, more like > a few dozens in each location, with locations being very remote and > having each their own cobbler instance, so I'm thinking about creating > a system entry in Cobbler for each and every host and guest I have. > Sure, this is good... > Obviously, they'll all inherit most of their configuration from a given > profile, but I'll need to override some settings here and there. > Hmm, yeah, take a look at the existing subprofile support, which may help. (The argument is --inherit and is covered in the manpage). > Matthias > > From mdehaan at redhat.com Thu Sep 27 15:40:00 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 27 Sep 2007 11:40:00 -0400 Subject: [et-mgmt-tools] WebUI issue In-Reply-To: <20070927131417.51880424@python3.es.egwn.lan> References: <46FAAA3D.3030704@redhat.com> <20070927130259.7a4e58fc@python3.es.egwn.lan> <20070927131417.51880424@python3.es.egwn.lan> Message-ID: <46FBCED0.1090808@redhat.com> Matthias Saou wrote: > Matthias Saou wrote : > > >> Michael DeHaan wrote : >> >> >>> This is fixed in the patch I linked earlier today. >>> Namely, there was a stray "#" in one of the templates. >>> If you want to pull from git and retest this should be happy now. >>> >>> Appreciate the help in looking things over. >>> >> The code I got from git when you wrote that this problem should have >> been fixed is still wrong. Apart from the s/random#/random/, you also >> moved the #end back to its own line, but "#end" isn't correct, it needs >> to be "#end if" for it to work apparently (I've just tested this). >> > > Replying to myself, but... > > With this now fixed, I get the ADD->SYSTEM page to display, but when I > click on the "random" link for the MAC address, the field gets filled > up with HTML code. > Yikes! :) Ok, I found the problem and some things that can cause it. I think we can conquer this by a minor bit of javascript, so I'll fix it before we push. Thanks! --Michael From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Thu Sep 27 15:47:39 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Thu, 27 Sep 2007 17:47:39 +0200 Subject: [et-mgmt-tools] Cobbler and Koan for _my_ needs In-Reply-To: <46FBCA49.6070500@redhat.com> References: <20070926185843.5f040191@python3.es.egwn.lan> <46FA99C5.6010806@redhat.com> <20070927105616.3d111c9f@python3.es.egwn.lan> <46FBBD7C.7010504@redhat.com> <20070927170306.1d5f20f7@python3.es.egwn.lan> <46FBCA49.6070500@redhat.com> Message-ID: <20070927174739.5a88ca81@python3.es.egwn.lan> Michael DeHaan wrote : > Hmm, this looks like leftover bits from when the name had to be a MAC. > Since the MAC > is a seperate field now, I'll see why it's not being used if it's provided. > > However, even so, you can still add the optional parameter --virt-name > to name your guest. > > koan --virt --system=AA:BB:CC:DD:EE:FF --server=cobbler.example.org --virt-name=blinky Indeed, this is fine as a temporary workaround :-) FWIW, using koan --display seems to imply that the MAC address isn't getting sent by Cobbler. > > I don't want to override anything on the koan command line. I'm more > > than fine with having everything in Cobbler! > > > > But in my case, I'd need my profile to be : > > - virt_path: /dev/data/$name,/dev/data/swap$name > > (I'm not sure if this can be done, I can live with overriding it in > > all system configurations, which is what I've done for now) > > - virt_ram: 512 > > > This would put them in two directories, which should be sufficient: > > --virt-path=/dev/data,/dev/data/swap/ I'm actually using unpartitioned LVM volumes for my guests :-) So the above uses the pre-created /dev/data/test and /dev/data/swaptest LVs. My tests have shown better performance that with files. I haven't tried not having the LVs pre-created and using "virt-path: '/dev/data/$name,/dev/data/swap$name'" with "virt_file_size: 4,2", but I suspect it won't work :-) > > Then the default system associated to be : > > - virt_path: <> > > - virt_ram: <> > > ...where I only change the name, hostname and MAC address. > > > > Here I'd like to be able to change virt_ram on a per-system basis, but > > it doesn't seem possible currently (or maybe I just need to "manually" > > add it to the system file?). > > > It's not like we couldn't add it, it just seems to go against the design > of having profiles > that define the requirements of what they are running. This is the same > reason you > can't request more virtual disk for a specific system -- the idea is > that you would > use inherited profiles to do this, and then just map the system to the > profile that it > was going to be assigned to. This will basically add one layer of configuration for very little to no gain in my case. I can live with that and do understand that the line must be hard to draw between inherited profiles and systems. It does seem like the WebUI doesn't currently know of inherited profiles, though. Thanks for all your help :-) Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.31 0.43 0.38 From mdehaan at redhat.com Thu Sep 27 15:51:29 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 27 Sep 2007 11:51:29 -0400 Subject: [et-mgmt-tools] Cobbler and Koan for _my_ needs In-Reply-To: <20070927174739.5a88ca81@python3.es.egwn.lan> References: <20070926185843.5f040191@python3.es.egwn.lan> <46FA99C5.6010806@redhat.com> <20070927105616.3d111c9f@python3.es.egwn.lan> <46FBBD7C.7010504@redhat.com> <20070927170306.1d5f20f7@python3.es.egwn.lan> <46FBCA49.6070500@redhat.com> <20070927174739.5a88ca81@python3.es.egwn.lan> Message-ID: <46FBD181.5090503@redhat.com> Matthias Saou wrote: > Michael DeHaan wrote : > > >> Hmm, this looks like leftover bits from when the name had to be a MAC. >> Since the MAC >> is a seperate field now, I'll see why it's not being used if it's provided. >> >> However, even so, you can still add the optional parameter --virt-name >> to name your guest. >> >> koan --virt --system=AA:BB:CC:DD:EE:FF --server=cobbler.example.org --virt-name=blinky >> > > Indeed, this is fine as a temporary workaround :-) > > FWIW, using koan --display seems to imply that the MAC address isn't > getting sent by Cobbler. > It is, it's just not being displayed :) > >>> I don't want to override anything on the koan command line. I'm more >>> than fine with having everything in Cobbler! >>> >>> But in my case, I'd need my profile to be : >>> - virt_path: /dev/data/$name,/dev/data/swap$name >>> (I'm not sure if this can be done, I can live with overriding it in >>> all system configurations, which is what I've done for now) >>> - virt_ram: 512 >>> >>> >> This would put them in two directories, which should be sufficient: >> >> --virt-path=/dev/data,/dev/data/swap/ >> > > I'm actually using unpartitioned LVM volumes for my guests :-) So the > above uses the pre-created /dev/data/test and /dev/data/swaptest LVs. > My tests have shown better performance that with files. > That works too. Specify two LVM volumes It was too early for my brain to notice the "/dev" in your path, apparently :) Names in LVM volumes are carved out based on the name of the virtual machine. > I haven't tried not having the LVs pre-created and using > "virt-path: '/dev/data/$name,/dev/data/swap$name'" with > "virt_file_size: 4,2", but I suspect it won't work :-) > > >>> Then the default system associated to be : >>> - virt_path: <> >>> - virt_ram: <> >>> ...where I only change the name, hostname and MAC address. >>> >>> Here I'd like to be able to change virt_ram on a per-system basis, but >>> it doesn't seem possible currently (or maybe I just need to "manually" >>> add it to the system file?). >>> >>> >> It's not like we couldn't add it, it just seems to go against the design >> of having profiles >> that define the requirements of what they are running. This is the same >> reason you >> can't request more virtual disk for a specific system -- the idea is >> that you would >> use inherited profiles to do this, and then just map the system to the >> profile that it >> was going to be assigned to. >> > > This will basically add one layer of configuration for very little to > no gain in my case. I can live with that and do understand that the > line must be hard to draw between inherited profiles and systems. > Kind of, yes. I'm trying to enforce the idiom of mapping systems to what they do. It's an important abstraction if you want to avoid managing a very large pool of systems with nothing in common, and especially important when you start scaling up into a very large number of systems (say, thousands). It makes things both predictable and repeatable. Inherited profiles are really there so that you can manage batches of profiles with some things in common (for instance, all db servers), but you might want to inherit from that profile to set settings that are site specific, or specific to a certain type of needy hardware. This makes interchanging hardware repeatable, because you contain that information in the profile, not in the individual (and non-reusable) system object. > It does seem like the WebUI doesn't currently know of inherited > profiles, though. > You are too observant :) This will likely come in a later release, when we figure out how to show those relationships best. Not a ton of people use them and I didn't want to add the complexity early on. > Thanks for all your help :-) > No problem. Thanks for all the good feedback and info on your environment. It's useful to hear from folks about this kind of thing. > Matthias > > From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Thu Sep 27 17:02:47 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Thu, 27 Sep 2007 19:02:47 +0200 Subject: [et-mgmt-tools] Cobbler and existing extra repos Message-ID: <20070927190247.3f8edc4d@python3.es.egwn.lan> Hi, I'm using Cobbler and a single kickstart file for both my EL5 i386 and EL5 x86_64 profiles. Now I've got the following problem : I'd like to include a "repo=" line in order to have anaconda use all of the available updates instead of having to run an update right after installation. I already have a yum repository of the updates on my LAN, so I don't want Cobbler to mirror anything. I just want it to add the repositories I give it through $yum_repo_stanza. Why don't I simply add the repo= line myself in the kickstart? Because anaconda doesn't replace the $basearch, and I want to share the same kickstart for i386 and x86_64. I can't seem to figure out how to add a simple name/baseurl as a repo to Cobbler. Is that possible at all? I've also tried the following, but it didn't work for some reason, and I don't really know how to debug it :-/ [...] url --url=$tree %include /tmp/repo [...] %pre #raw # Since the "repo" lines don't understand $basearch, hack that here basearch=$(uname -i) for repo in updates custom; do echo "repo --name=$repo --baseurl=http://myserver/redhat/el5Server/$basearch/$repo/" >> /tmp/repo done #end raw The --available-as import option made it possible to not mirror the base files locally (I've mounted them from my main yum server through nfs). Any ideas about what the easiest is to achieve what I need? Another solution would be for Cobbler to expose the basearch so that it can be used as $basearch in the kickstart and get replaced by the cheetah parsing. Currently I only see $arch in the Wiki, but it expands to "x86" and not "i386". Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.32 0.32 0.35 From mdehaan at redhat.com Thu Sep 27 17:00:15 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Thu, 27 Sep 2007 13:00:15 -0400 Subject: [et-mgmt-tools] Cobbler and existing extra repos In-Reply-To: <20070927190247.3f8edc4d@python3.es.egwn.lan> References: <20070927190247.3f8edc4d@python3.es.egwn.lan> Message-ID: <46FBE19F.1070607@redhat.com> Matthias Saou wrote: > Hi, > > I'm using Cobbler and a single kickstart file for both my EL5 i386 and > EL5 x86_64 profiles. Now I've got the following problem : I'd like to > include a "repo=" line in order to have anaconda use all of the > available updates instead of having to run an update right after > installation. > Certaintly... > I already have a yum repository of the updates on my LAN, so I don't > want Cobbler to mirror anything. I just want it to add the repositories > I give it through $yum_repo_stanza. > > Why don't I simply add the repo= line myself in the kickstart? > Because anaconda doesn't replace the $basearch, and I want to share the > same kickstart for i386 and x86_64. > You can use kickstart templating with an #if statement based on $arch which is settable in each distro object. > Another solution would be for Cobbler to expose the basearch so that it > can be used as $basearch in the kickstart and get replaced by the > cheetah parsing. Currently I only see $arch in the Wiki, but it expands > to "x86" and not "i386". > Yep. All you need is an if statement as these mean the same thing. > Matthias > > From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Thu Sep 27 17:16:00 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Thu, 27 Sep 2007 19:16:00 +0200 Subject: [et-mgmt-tools] Cobbler and existing extra repos In-Reply-To: <46FBE19F.1070607@redhat.com> References: <20070927190247.3f8edc4d@python3.es.egwn.lan> <46FBE19F.1070607@redhat.com> Message-ID: <20070927191600.26506fed@python3.es.egwn.lan> Michael DeHaan wrote : > You can use kickstart templating with an #if statement based on $arch > which is settable > in each distro object. D'oh! Thanks, Michael :-D Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.43 0.41 0.37 From rjones at redhat.com Thu Sep 27 18:07:00 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Thu, 27 Sep 2007 19:07:00 +0100 Subject: [et-mgmt-tools] virt-df (a 'df' tool for virtual domains) Message-ID: <46FBF144.2010505@redhat.com> Brave (or foolhardy?) souls may want to try out the almost-working version of 'virt-df' that I wrote: hg clone http://hg.et.redhat.com/virt/applications/virt-top--devel Usage is similar to ordinary 'df'. It understands the -h (human-readable) and -i (show inodes) options. You need to run it on the dom0, it won't work remotely. # virt-df.opt -c qemu:///system debian32kvm /dev/hda (3.9 GiB) /var/lib/xen/images/debian32kvm.img hda1 3662792 990276 2672516 Linux ext2/3 hda5 232908 Linux swap [The numbers are 1K blocks total, used, available] # virt-df.opt -c qemu:///system -h debian32kvm /dev/hda (3.9 GiB) /var/lib/xen/images/debian32kvm.img hda1 3.5 GiB 967.1 MiB 2.5 GiB Linux ext2/3 hda5 227.4 MiB Linux swap From the README file: virt-df is a 'df' tool for printing out the used and available disk space in all active and inactive domains. Without this tool you would need to log in to each domain individually or set up monitoring. It is only a proof-of-concept. Please bare in mind the following limitations when using this tool: (1) It does not work over remote connections. Part of the reason why I wrote virt-df was to get an idea of how the remote storage API for libvirt might look. (2) It only understands a limited set of partition types. Assuming that the files and partitions that we get back from libvirt / Xen correspond to block devices in the guests, we can go some way towards manually parsing those partitions to find out what they contain. We can read the MBR, EBR, superblocks and so on. However that's a lot of parsing work, and currently there is no library which understands a wide range of partition schemes and filesystem types (not even libparted which doesn't support LVM yet). The Linux kernel does support that, but there's not really any good way to access that work. The current implementation uses a hand-coded parser which understands some simple formats (MBR, EBR, ext2/3, not LVM yet but coming soon). In future we should use something like libparted. (3) The statistics you get are delayed. The real state of, for example, an ext2 filesystem is only stored in the memory of the guest's kernel. The ext2 superblock contains some meta-information about blocks used and free, but this superblock is not up to date. In fact the guest kernel may not update it even on a 'sync', not until the filesystem is unmounted. Some operations do appear to write the superblock, for example fsync(2) [that is my reading of the ext2/3 source code at least]. $ wc -l virt-df/*.ml 84 virt-df/virt_df_ext2.ml 24 virt-df/virt_df_linux_swap.ml 22 virt-df/virt_df_lvm2.ml 5 virt-df/virt_df_main.ml 460 virt-df/virt_df.ml 595 total 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 levon at movementarian.org Thu Sep 27 18:36:35 2007 From: levon at movementarian.org (John Levon) Date: Thu, 27 Sep 2007 19:36:35 +0100 Subject: [et-mgmt-tools] Re: [Libvir] virt-df (a 'df' tool for virtual domains) In-Reply-To: <46FBF144.2010505@redhat.com> References: <46FBF144.2010505@redhat.com> Message-ID: <20070927183635.GA10845@totally.trollied.org.uk> On Thu, Sep 27, 2007 at 07:07:00PM +0100, Richard W.M. Jones wrote: > (2) It only understands a limited set of partition types. Assuming that > the files and partitions that we get back from libvirt / Xen correspond > to block devices in the guests, we can go some way towards manually > parsing those partitions to find out what they contain. We can read the > MBR, EBR, superblocks and so on. However that's a lot of parsing work, > and currently there is no library which understands a wide range of > partition schemes and filesystem types (not even libparted which doesn't > support LVM yet). The Linux kernel does support that, but there's not > really any good way to access that work. Note that extending libfsimage would be the way to go for the latter part of this (understanding file systems), I think. regards, john From rjones at redhat.com Thu Sep 27 18:46:53 2007 From: rjones at redhat.com (Richard W.M. Jones) Date: Thu, 27 Sep 2007 19:46:53 +0100 Subject: [et-mgmt-tools] Re: [Libvir] virt-df (a 'df' tool for virtual domains) In-Reply-To: <20070927183635.GA10845@totally.trollied.org.uk> References: <46FBF144.2010505@redhat.com> <20070927183635.GA10845@totally.trollied.org.uk> Message-ID: <46FBFA9D.1000605@redhat.com> John Levon wrote: > On Thu, Sep 27, 2007 at 07:07:00PM +0100, Richard W.M. Jones wrote: > >> (2) It only understands a limited set of partition types. Assuming that >> the files and partitions that we get back from libvirt / Xen correspond >> to block devices in the guests, we can go some way towards manually >> parsing those partitions to find out what they contain. We can read the >> MBR, EBR, superblocks and so on. However that's a lot of parsing work, >> and currently there is no library which understands a wide range of >> partition schemes and filesystem types (not even libparted which doesn't >> support LVM yet). The Linux kernel does support that, but there's not >> really any good way to access that work. > > Note that extending libfsimage would be the way to go for the latter > part of this (understanding file systems), I think. Yes ... There are (at least) four separate reimplementations of filesystem / partition parsing code: - fdisk - libparted - grub (from which libfsimage seems to be derived) - Linux kernel Unfortunately libparted & GRUB don't understand LVM which is kind of a problem because everyone's using LVM nowadays. (Well, everyone in the Linux world, obviously different on Solaris). There is some work that I'm aware of to "librarify" LVM and to get LVM support into libparted. Having said all that, the hand-written code I wrote deals with MBRs, EBRs and ext2/3 and it's under 600 lines of code, and even with LVM support I can't see it going over 1,500 lines in total. The big question is what sort of API could libvirt export? For the general provisioning scenario, perhaps it would be something like this: https://www.redhat.com/archives/libvir-list/2007-September/thread.html#00119 Reading and understanding the contents of block devices OTOH is way beyond what libvirt can or ought to provide. Instead it can export a way to snoop on the contents of the block devices (simple operations like: read the size, seek to an offset, read some bytes); then we could extend libparted to abstract away the Unix syscalls so it can work on the libvirt "virtual block device". 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 berrange at redhat.com Thu Sep 27 19:05:10 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 27 Sep 2007 20:05:10 +0100 Subject: [et-mgmt-tools] Re: [Libvir] virt-df (a 'df' tool for virtual domains) In-Reply-To: <46FBFA9D.1000605@redhat.com> References: <46FBF144.2010505@redhat.com> <20070927183635.GA10845@totally.trollied.org.uk> <46FBFA9D.1000605@redhat.com> Message-ID: <20070927190510.GO17433@redhat.com> On Thu, Sep 27, 2007 at 07:46:53PM +0100, Richard W.M. Jones wrote: > John Levon wrote: > >On Thu, Sep 27, 2007 at 07:07:00PM +0100, Richard W.M. Jones wrote: > > > >>(2) It only understands a limited set of partition types. Assuming that > >>the files and partitions that we get back from libvirt / Xen correspond > >>to block devices in the guests, we can go some way towards manually > >>parsing those partitions to find out what they contain. We can read the > >>MBR, EBR, superblocks and so on. However that's a lot of parsing work, > >>and currently there is no library which understands a wide range of > >>partition schemes and filesystem types (not even libparted which doesn't > >>support LVM yet). The Linux kernel does support that, but there's not > >>really any good way to access that work. > > > >Note that extending libfsimage would be the way to go for the latter > >part of this (understanding file systems), I think. > > Yes ... There are (at least) four separate reimplementations of > filesystem / partition parsing code: > > - fdisk > - libparted > - grub (from which libfsimage seems to be derived) > - Linux kernel > > Unfortunately libparted & GRUB don't understand LVM which is kind of a > problem because everyone's using LVM nowadays. (Well, everyone in the > Linux world, obviously different on Solaris). > > There is some work that I'm aware of to "librarify" LVM and to get LVM > support into libparted. > > Having said all that, the hand-written code I wrote deals with MBRs, > EBRs and ext2/3 and it's under 600 lines of code, and even with LVM > support I can't see it going over 1,500 lines in total. > > The big question is what sort of API could libvirt export? For the > general provisioning scenario, perhaps it would be something like this: > > https://www.redhat.com/archives/libvir-list/2007-September/thread.html#00119 There is the prior discussion on storage requirements too: http://www.redhat.com/archives/libvir-list/2007-February/msg00010.html http://www.redhat.com/archives/libvir-list/2007-February/msg00057.html None of these threads really considered the case of iSCSI, or SAN storage in sufficient detail yet, or the latest SCSI storage adapter virtaulization capabilities that are now appearing. We really need to get a better conceptual model of what we want to represent & enable before diving into APIs. There are two clearly different sets of APIs we're talking about here, but they're getting mixed up. There are APIs for managing storage volumes, and then there are the APIs for dealing with filesystems within them. I could see the libfsimage being useful, except that it is GPL licensed at the moment. Getting free space stats is handy for monitoring. Being able to ask whether there is a valid MBR, and/or Grub config is also handy to tell whether an install was succcess - virt-install for example has a hack to read the first few bytes and look for an MBR to check install status. 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 Thu Sep 27 20:15:16 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Thu, 27 Sep 2007 21:15:16 +0100 Subject: [et-mgmt-tools] Re: [Libvir] virt-df (a 'df' tool for virtual domains) In-Reply-To: <20070927190510.GO17433@redhat.com> References: <46FBF144.2010505@redhat.com> <20070927183635.GA10845@totally.trollied.org.uk> <46FBFA9D.1000605@redhat.com> <20070927190510.GO17433@redhat.com> Message-ID: <20070927201516.GP17433@redhat.com> On Thu, Sep 27, 2007 at 08:05:10PM +0100, Daniel P. Berrange wrote: > On Thu, Sep 27, 2007 at 07:46:53PM +0100, Richard W.M. Jones wrote: > > John Levon wrote: > > >On Thu, Sep 27, 2007 at 07:07:00PM +0100, Richard W.M. Jones wrote: > > > > > >>(2) It only understands a limited set of partition types. Assuming that > > >>the files and partitions that we get back from libvirt / Xen correspond > > >>to block devices in the guests, we can go some way towards manually > > >>parsing those partitions to find out what they contain. We can read the > > >>MBR, EBR, superblocks and so on. However that's a lot of parsing work, > > >>and currently there is no library which understands a wide range of > > >>partition schemes and filesystem types (not even libparted which doesn't > > >>support LVM yet). The Linux kernel does support that, but there's not > > >>really any good way to access that work. > > > > > >Note that extending libfsimage would be the way to go for the latter > > >part of this (understanding file systems), I think. > > > > Yes ... There are (at least) four separate reimplementations of > > filesystem / partition parsing code: > > > > - fdisk > > - libparted > > - grub (from which libfsimage seems to be derived) > > - Linux kernel > > > > Unfortunately libparted & GRUB don't understand LVM which is kind of a > > problem because everyone's using LVM nowadays. (Well, everyone in the > > Linux world, obviously different on Solaris). > > > > There is some work that I'm aware of to "librarify" LVM and to get LVM > > support into libparted. > > > > Having said all that, the hand-written code I wrote deals with MBRs, > > EBRs and ext2/3 and it's under 600 lines of code, and even with LVM > > support I can't see it going over 1,500 lines in total. > > > > The big question is what sort of API could libvirt export? For the > > general provisioning scenario, perhaps it would be something like this: > > > > https://www.redhat.com/archives/libvir-list/2007-September/thread.html#00119 > > There is the prior discussion on storage requirements too: > > http://www.redhat.com/archives/libvir-list/2007-February/msg00010.html > http://www.redhat.com/archives/libvir-list/2007-February/msg00057.html > > None of these threads really considered the case of iSCSI, or SAN storage > in sufficient detail yet, or the latest SCSI storage adapter virtaulization > capabilities that are now appearing. We really need to get a better conceptual > model of what we want to represent & enable before diving into APIs. > > There are two clearly different sets of APIs we're talking about here, but > they're getting mixed up. There are APIs for managing storage volumes, and > then there are the APIs for dealing with filesystems within them. > > I could see the libfsimage being useful, except that it is GPL licensed at > the moment. Getting free space stats is handy for monitoring. Being able to > ask whether there is a valid MBR, and/or Grub config is also handy to tell > whether an install was succcess - virt-install for example has a hack to > read the first few bytes and look for an MBR to check install status. John pointed out to me on IRC that its only the plugins which are GPL. The core library is BSD licensed. The plugins are dl-open() by the main lib - not yet sure what that implies license-wise though... Depending on the way we do the storage APIs this may or may not be an issue. ie, only the libvirt.so needs to be LGPL. If the daemon itself becomes GPL by virtue of loading a GPL library that's not neccessarily a terminal problem if we decide that is worth it, since no end user apps link to the daemon. 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 Guillaume.Rousse at inria.fr Thu Sep 27 22:03:36 2007 From: Guillaume.Rousse at inria.fr (Guillaume Rousse) Date: Fri, 28 Sep 2007 00:03:36 +0200 Subject: [et-mgmt-tools] Accentued caracters display issue in non UTF8 environment Message-ID: <46FC28B8.4090404@inria.fr> virt-manager doesn't display french accentuated characters correctly in non-UTF environment (fr_FR). See http://qa.mandriva.com/show_bug.cgi?id=34184 for details. -- Guillaume Rousse Moyens Informatiques - INRIA Futurs Tel: 01 69 35 69 62 From berrange at redhat.com Thu Sep 27 23:45:52 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Fri, 28 Sep 2007 00:45:52 +0100 Subject: [et-mgmt-tools] Accentued caracters display issue in non UTF8 environment In-Reply-To: <46FC28B8.4090404@inria.fr> References: <46FC28B8.4090404@inria.fr> Message-ID: <20070927234552.GA6258@redhat.com> On Fri, Sep 28, 2007 at 12:03:36AM +0200, Guillaume Rousse wrote: > virt-manager doesn't display french accentuated characters correctly in > non-UTF environment (fr_FR). See > http://qa.mandriva.com/show_bug.cgi?id=34184 for details. All the .po files are created with a UTF-8 charset, so your locale is not UTF-8 then that would certainly cause messed up accentuated characters. To be honest I don't know what todo here. Out of all the .mo files I have installed in /usr/share/locale/fr/LC_MESSAGES/ 170 are in UTF-8, and only 30 in iso8859-1, so virt-manager is following most common behavuour of other apps which are translated AFAICT. My preference is to say just use UTF-8 locales, unless someone has better suggestions. 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 wright at imageworks.com Fri Sep 28 00:17:09 2007 From: wright at imageworks.com (Peter Wright) Date: Thu, 27 Sep 2007 17:17:09 -0700 Subject: [et-mgmt-tools] koan create xen domU MAC issue Message-ID: <46FC4805.4000501@imageworks.com> hi all - has anyone had any issues with MAC addresses when creating a new xen domU instance via koan? here is what i'm running into: create host in cobbler with this MAC addy: 00:9c:5A:2c:50:8F run "cobbler report"...looks good, run "cobbler sync"...no problems. run this koan line: koan --virt --server=maa-master --system=maa-xen2 --virt-path=/var/xen --virt-name=xen2 cat /etc/xen/xen2: [root at maa12 xen]# cat /etc/xen/xen2 disk = [ 'file:/var/xen/xen2,xvda,w', ] vif = [ 'mac=00:16:3e:2e:27:5e, bridge=xenbr0', ] so it looks like the MAC i'm assigning in cobbler is not being assigned to my xen instance. is this expected, or am i missing something here? thanks! -pete -- Peter Wright Systems Engineer Sony Pictures Imageworks wright at imageworks.com www.imageworks.com From herrold at owlriver.com Fri Sep 28 02:52:14 2007 From: herrold at owlriver.com (R P Herrold) Date: Thu, 27 Sep 2007 22:52:14 -0400 (EDT) Subject: [et-mgmt-tools] Re: et-mgnt] Cobbler and Koan for _my_ needs In-Reply-To: <20070927170306.1d5f20f7@python3.es.egwn.lan> References: <20070926185843.5f040191@python3.es.egwn.lan> <46FA99C5.6010806@redhat.com> <20070927105616.3d111c9f@python3.es.egwn.lan> <46FBBD7C.7010504@redhat.com> <20070927170306.1d5f20f7@python3.es.egwn.lan> Message-ID: On Thu, 27 Sep 2007, Matthias Saou wrote: > Hmmmm, indeed, this works, but this makes things uglier for me. > Renaming my system from 'test' to '00:16:3e:01:ec:2e' in Cobbler and > using --system=00:16:3e:01:ec:2e instead of --system=test makes the > virtual guest get the intended the MAC address, but : > When I run 'xm list' and see a bunch of host names, it's much quicker > for me to stop/restart/connect-to any of the guests than with having a > list of 00_16_3e_xx_xx_xx entries :-/ Same for the WebUI... > > So it seems to me that there is too much logic on the Koan side, which > makes it impossible to use a friendly name and have it use the MAC > address configured in the Cobbler. It is not magic to do a look up from a hard name (a hash, actually) to common name. Is an RFE in order here, to add a backend transation store, and minimal maintenance tools to manage it? -- sqlite and just a few lines of code, and one is done. - or - add perl-DBI perlish hooks, and one gets whatever backing store is locally convenient. -- Russ Herrold From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Fri Sep 28 08:40:40 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Fri, 28 Sep 2007 10:40:40 +0200 Subject: [et-mgmt-tools] koan create xen domU MAC issue In-Reply-To: <46FC4805.4000501@imageworks.com> References: <46FC4805.4000501@imageworks.com> Message-ID: <20070928104040.09d69dbd@python3.es.egwn.lan> Peter Wright wrote : > hi all - has anyone had any issues with MAC addresses when creating a > new xen domU instance via koan? here is what i'm running into: > > > create host in cobbler with this MAC addy: 00:9c:5A:2c:50:8F > > run "cobbler report"...looks good, run "cobbler sync"...no problems. > > run this koan line: > koan --virt --server=maa-master --system=maa-xen2 --virt-path=/var/xen > --virt-name=xen2 > > cat /etc/xen/xen2: > [root at maa12 xen]# cat /etc/xen/xen2 > > disk = [ 'file:/var/xen/xen2,xvda,w', ] > vif = [ 'mac=00:16:3e:2e:27:5e, bridge=xenbr0', ] > > > > so it looks like the MAC i'm assigning in cobbler is not being assigned > to my xen instance. is this expected, or am i missing something here? I reported the same issue here, and Michael wrote that it was probably a leftover bug from when names were MAC addresses. Temporary workaround : Rename your Cobbler system from "maa-xen2" to "00:9c:5a:2c:50:8f" and it should work. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.36 0.46 0.40 From thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net Fri Sep 28 11:53:44 2007 From: thias at spam.spam.spam.spam.spam.spam.spam.egg.and.spam.freshrpms.net (Matthias Saou) Date: Fri, 28 Sep 2007 13:53:44 +0200 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <46F9902A.8060205@redhat.com> References: <46F9902A.8060205@redhat.com> Message-ID: <20070928135344.1eaf852d@python3.es.egwn.lan> Michael DeHaan wrote : > So ... I'm uploading the 0.6.2 RPM's today for folks to have a chance to > try things out before pushing to the mirrors on Friday. > > http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm > http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm These versions should have fixed this bug, right? https://hosted.fedoraproject.org/projects/koan/ticket/5 I'm asking, because I've just tried again as follows : # koan -s cobbler -y test -v -V test -b xenbr1 But the /etc/xen/test file contains : vif = [ 'mac=00:16:3e:34:61:3e, bridge=xenbr0', ] So it seems that something is still going wrong. Matthias -- Clean custom Red Hat Linux rpm packages : http://freshrpms.net/ Fedora release 7 (Moonshine) - Linux kernel 2.6.22.6-81.fc7 Load : 0.65 0.47 0.44 From mdehaan at redhat.com Fri Sep 28 14:29:39 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 28 Sep 2007 10:29:39 -0400 Subject: [et-mgmt-tools] koan create xen domU MAC issue In-Reply-To: <46FC4805.4000501@imageworks.com> References: <46FC4805.4000501@imageworks.com> Message-ID: <46FD0FD3.2050606@redhat.com> Peter Wright wrote: > hi all - has anyone had any issues with MAC addresses when creating a > new xen domU instance via koan? here is what i'm running into: > > > run this koan line: > koan --virt --server=maa-master --system=maa-xen2 --virt-path=/var/xen > --virt-name=xen2 > This was reported yesterday on the list, in fact :) If you name the system after the mac address (the old way of doing it) it seems to work, otherwise that's something I need to fix for the next release. From mdehaan at redhat.com Fri Sep 28 14:31:10 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 28 Sep 2007 10:31:10 -0400 Subject: [et-mgmt-tools] Re: et-mgnt] Cobbler and Koan for _my_ needs In-Reply-To: References: <20070926185843.5f040191@python3.es.egwn.lan> <46FA99C5.6010806@redhat.com> <20070927105616.3d111c9f@python3.es.egwn.lan> <46FBBD7C.7010504@redhat.com> <20070927170306.1d5f20f7@python3.es.egwn.lan> Message-ID: <46FD102E.1040502@redhat.com> R P Herrold wrote: > On Thu, 27 Sep 2007, Matthias Saou wrote: > >> Hmmmm, indeed, this works, but this makes things uglier for me. >> Renaming my system from 'test' to '00:16:3e:01:ec:2e' in Cobbler and >> using --system=00:16:3e:01:ec:2e instead of --system=test makes the >> virtual guest get the intended the MAC address, but : > >> When I run 'xm list' and see a bunch of host names, it's much quicker >> for me to stop/restart/connect-to any of the guests than with having a >> list of 00_16_3e_xx_xx_xx entries :-/ Same for the WebUI... >> >> So it seems to me that there is too much logic on the Koan side, which >> makes it impossible to use a friendly name and have it use the MAC >> address configured in the Cobbler. > > It is not magic to do a look up from a hard name (a hash, actually) to > common name. Is an RFE in order here, to add a backend transation > store, and minimal maintenance tools to manage it? -- sqlite and just > a few lines of code, and one is done. - or - add perl-DBI perlish > hooks, and one gets whatever backing store is locally convenient. > This really has nothing to do with databases. This has to do with checking a different field name than what is currently checked. It'll be in the next release. --Michael From mdehaan at redhat.com Fri Sep 28 14:35:20 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 28 Sep 2007 10:35:20 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler 0.6.2 (pre) release In-Reply-To: <20070928135344.1eaf852d@python3.es.egwn.lan> References: <46F9902A.8060205@redhat.com> <20070928135344.1eaf852d@python3.es.egwn.lan> Message-ID: <46FD1128.40300@redhat.com> Matthias Saou wrote: > Michael DeHaan wrote : > > >> So ... I'm uploading the 0.6.2 RPM's today for folks to have a chance to >> try things out before pushing to the mirrors on Friday. >> >> http://cobbler.et.redhat.com/download/cobbler-0.6.2-1.src.rpm >> http://cobbler.et.redhat.com/download/koan-0.6.2-1.src.rpm >> > > These versions should have fixed this bug, right? > https://hosted.fedoraproject.org/projects/koan/ticket/5 > > I'm asking, because I've just tried again as follows : > > # koan -s cobbler -y test -v -V test -b xenbr1 > I should have never put the short args in :) > But the /etc/xen/test file contains : > vif = [ 'mac=00:16:3e:34:61:3e, bridge=xenbr0', ] > > So it seems that something is still going wrong. > It may be that your version of libvirt doesn't support that feature. The code in question defaults to "old-style" network setup when it can't successfully add the interface. You may want to upgrade your version of virtinst. > Matthias > > From wright at imageworks.com Fri Sep 28 15:51:07 2007 From: wright at imageworks.com (Peter Wright) Date: Fri, 28 Sep 2007 08:51:07 -0700 Subject: [et-mgmt-tools] koan create xen domU MAC issue In-Reply-To: <46FD0FD3.2050606@redhat.com> References: <46FC4805.4000501@imageworks.com> <46FD0FD3.2050606@redhat.com> Message-ID: <46FD22EB.6020600@imageworks.com> Michael DeHaan wrote: > > Peter Wright wrote: > > hi all - has anyone had any issues with MAC addresses when creating a > > new xen domU instance via koan? here is what i'm running into: > > > > > > run this koan line: > > koan --virt --server=maa-master --system=maa-xen2 --virt-path=/var/xen > > --virt-name=xen2 > > > > This was reported yesterday on the list, in fact :) > > If you name the system after the mac address (the old way of doing it) > it seems to work, otherwise that's something I need to fix for the next > release. > thanks all! - sorry for the noise, i didn't do my own due diligence of actually reading my email to see if someone else had run into this before complaining :) -pete -- Peter Wright Systems Engineer Sony Pictures Imageworks wright at imageworks.com www.imageworks.com From crobinso at redhat.com Fri Sep 28 17:08:19 2007 From: crobinso at redhat.com (Cole Robinson) Date: Fri, 28 Sep 2007 13:08:19 -0400 Subject: [et-mgmt-tools] [PATCH] reworked virtinst validation testing Message-ID: <46FD3503.3020008@redhat.com> Hi all, A couple months back I sent some virtinst validation tests to the list but the patch was never applied. I reworked it against the current tip and improved it a bit, and here is the result! Thanks, Cole -- Cole Robinson crobinso at redhat.com -------------- next part -------------- A non-text attachment was scrubbed... Name: virtinst-validation-testing.patch Type: text/x-patch Size: 12147 bytes Desc: not available URL: From crobinso at redhat.com Fri Sep 28 17:16:27 2007 From: crobinso at redhat.com (Cole Robinson) Date: Fri, 28 Sep 2007 13:16:27 -0400 Subject: [et-mgmt-tools] [PATCH] reworked virtinst validation fixes Message-ID: <46FD36EB.2080402@redhat.com> Hi all, This patch fixes some validation errors in virtinst uncovered by the tests I just sent. I sent this a couple months ago but it must have just fallen through the cracks, so I reworked it against the current tip. This stuff is pretty simple and handy for useful error reporting. Thanks, Cole -- Cole Robinson crobinso at redhat.com -------------- next part -------------- A non-text attachment was scrubbed... Name: virtinst-validation-fixes.patch Type: text/x-patch Size: 9442 bytes Desc: not available URL: From berrange at redhat.com Fri Sep 28 19:30:00 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Fri, 28 Sep 2007 20:30:00 +0100 Subject: [et-mgmt-tools] [PATCH] reworked virtinst validation fixes In-Reply-To: <46FD36EB.2080402@redhat.com> References: <46FD36EB.2080402@redhat.com> Message-ID: <20070928193000.GC14112@redhat.com> On Fri, Sep 28, 2007 at 01:16:27PM -0400, Cole Robinson wrote: > Hi all, > > This patch fixes some validation errors in virtinst uncovered by the > tests I just sent. I sent this a couple months ago but it must have just > fallen through the cracks, so I reworked it against the current tip. > This stuff is pretty simple and handy for useful error reporting. Looks ok, except for this part: - if path != None: - self.path = os.path.abspath(path) - else: - self.path = None - type = VirtualDisk.TYPE_FILE # Arbitrary choice but avoids the null-path null-type case - + self.path = path + + if __builtin__.type(self.path) is not __builtin__.type("string"): + raise ValueError, _("The disk path must be a string.") + self.path = os.path.abspath(self.path) + And a few places following it which don't cope with fact that self.path is allowed to be None. Path is allowed to be None for CDROMs so we can setup a disk whichout any media inserted. BTW, can you fix your Thunderbird so it either sends patches as text/plain, or just inlines them. http://lwn.net/Articles/249669/ Using text/x-patch as a type makes quoting & replying harder. Regards, Dan. -- |=- Red Hat, Engineering, Emerging Technologies, Boston. +1 978 392 2496 -=| |=- Perl modules: http://search.cpan.org/~danberr/ -=| |=- Projects: http://freshmeat.net/~danielpb/ -=| |=- GnuPG: 7D3B9505 F3C9 553F A1DA 4AC2 5648 23C1 B3DF F742 7D3B 9505 -=| From mdehaan at redhat.com Fri Sep 28 21:03:37 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 28 Sep 2007 17:03:37 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Func 0.0.12 Message-ID: <46FD6C29.5040809@redhat.com> Hi, New software release time! Yay! Lots of us have been thinking about and working on a system to help make control of a very large number of remote machines easier. To save myself the trouble of writing the release-announcement, I'll just borrow from our meticulously crafted web page and then let you read that :) https://hosted.fedoraproject.org/projects/func/ * Have you ever tried to manage a large number of systems with SSH? Have you wanted a better way? * Have you wanted a way to audit all of your remote commands on all of your systems? * Tired of writing shell scripts and parsing command output? * Are you fed up with CIM, WBEM, and complicated systems that prevent you from doing /real/ work? * Well have we got a solution for you. It's Func. There are now src RPM and tarball downloads of the first version up on the Wiki page under "FuncReleases". You can also pull from git if you are so inclined. It's definitely an alpha/beta kind of thing as this point, so it's a great time to start checking it, getting involved, and contributing your own ideas, modules, scripts, and so on. For more information: * https://hosted.fedoraproject.org/projects/func/ * https://www.redhat.com/mailman/listinfo/func-list * #func on irc.freenode.net As usual comments and questions are welcome. However, please use the func-mailing list and not this one if you can remember. Future announcements will go there, as we don't want to flood this list with development topics on func. Func is brought to you by... James Bowes Michael DeHaan Greg DeKoenigsberg Adrian Likins Robin Norwood Seth Vidal Jesus Rodriguez (hope I'm not leaving anyone out) and soon to be ... you :) Thanks! --Michael From dlutter at redhat.com Fri Sep 28 22:16:18 2007 From: dlutter at redhat.com (David Lutterkort) Date: Fri, 28 Sep 2007 15:16:18 -0700 Subject: [et-mgmt-tools] [ANNOUNCE] Func 0.0.12 In-Reply-To: <46FD6C29.5040809@redhat.com> References: <46FD6C29.5040809@redhat.com> Message-ID: <1191017778.22064.9.camel@localhost.localdomain> On Fri, 2007-09-28 at 17:03 -0400, Michael DeHaan wrote: > https://hosted.fedoraproject.org/projects/func/ I'm a little confused as to the intended use/target audience. As a tool for casual settings, this seems fine, though > * Have you ever tried to manage a large number of systems with SSH? > Have you wanted a better way? Anybody who manages their machines that way (and thinks they are truly managing them) needs to go back to sysadmin school. (Chapter 31 in Limoncelli is probably a good starting point, though the rest of the book is a very good read, too) > * Have you wanted a way to audit all of your remote commands on all > of your systems? > * Tired of writing shell scripts and parsing command output? Isn't that all stuff that a config mgmt system does ? > * Are you fed up with CIM, WBEM, and complicated systems that > prevent you from doing /real/ work? There are config mgmt systems that don't use CIM; puppet[1] would be my recommendation. David [1] http://puppet.reductivelabs.com/ From mdehaan at redhat.com Fri Sep 28 22:15:47 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 28 Sep 2007 18:15:47 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Cobbler and koan 0.6.2-2 Message-ID: <46FD7D13.9080106@redhat.com> Hi folks, Thanks very much to everyone who helped test the Web UI this week! I've incorporated everyone's testing reports into a new release, which I have pushed out to the Fedora and EPEL Testing mirrors. You can also get the RPMs from http://cobbler.et.redhat.com/download/. As I stated before, the WebUI is still a beta at this point, so please test away at it, and we'll continue to incorporate your feedback. I've been thinking about cobbler/koan 0.6.3, and at minimum, here's what I'm going to be looking at adding for the next release: cobbler -- paginate the systems list page in the Web UI so you don't have to scroll when you have 1000+ systems :) cobbler -- support creation of inherited profiles from within the WebUI koan -- support creation of multiple network interfaces as opposed to just one (a couple of folks have asked for this) koan -- if name is not a MAC address, but a MAC is specified in --mac, fix the bug that makes the virtual MAC address be random. potentially look at a few items from the Trac Misc RFE page -- https://hosted.fedoraproject.org/projects/cobbler/report/1 look again at developing a sqlite-backed serializer, mainly for performance reasons with very large numbers of systems to serialize possibly finally move /var/lib/cobbler/settings to /etc/cobbler/settings :) As usual, if anyone has any other features they would like to see, let me know -- and I'll takes patches too :) --Michael From mdehaan at redhat.com Fri Sep 28 23:12:37 2007 From: mdehaan at redhat.com (Michael DeHaan) Date: Fri, 28 Sep 2007 19:12:37 -0400 Subject: [et-mgmt-tools] [ANNOUNCE] Func 0.0.12 In-Reply-To: <1191017778.22064.9.camel@localhost.localdomain> References: <46FD6C29.5040809@redhat.com> <1191017778.22064.9.camel@localhost.localdomain> Message-ID: <46FD8A65.3000206@redhat.com> >> * Have you ever tried to manage a large number of systems with SSH? >> Have you wanted a better way? >> > > Anybody who manages their machines that way (and thinks they are truly > managing them) needs to go back to sysadmin school. As I'm running a bit late and need to get going, I'll leave that for someone else to respond to in detail -- but there are a lot of extremely intelligent sysadmins running thousands of machine configurations out there, many of which who are on these lists. A couple of them have been working with us to build this. They've mentioned they've wanted better tools around key exchange, and ways to perform one-offs and for data gathering -- things that don't fit well into the cfengine/puppet kind of models. Lots of us who decided to build this think similar things, and at least four of us have said "I thought about building this at one time or another and never did". So yeah, we think it has uses. Configuration Management tools are great at declaring the way a system should look, and are a great hammer, but sometimes you need an allen wrench. You can use both. They are not competing. Anyhow, this is part of the reason we're not putting func explicitly on a systems management email list ... Func has uses other than systems management, for other apps that want a generic key exchange mechanism and modular XMLRPC framework. That's what we're building. We expect all sorts of Fedora apps to start using the infrastructure, because it's small, simple, and nice and Pythonic. Does it have systems management uses? Certaintly. Is it a Puppet replacement? No. --Michael From dlutter at redhat.com Sat Sep 29 00:01:22 2007 From: dlutter at redhat.com (David Lutterkort) Date: Fri, 28 Sep 2007 17:01:22 -0700 Subject: [et-mgmt-tools] [ANNOUNCE] Func 0.0.12 In-Reply-To: <46FD8A65.3000206@redhat.com> References: <46FD6C29.5040809@redhat.com> <1191017778.22064.9.camel@localhost.localdomain> <46FD8A65.3000206@redhat.com> Message-ID: <1191024082.22064.25.camel@localhost.localdomain> On Fri, 2007-09-28 at 19:12 -0400, Michael DeHaan wrote: > >> * Have you ever tried to manage a large number of systems with SSH? > >> Have you wanted a better way? > >> > > > > Anybody who manages their machines that way (and thinks they are truly > > managing them) needs to go back to sysadmin school. > > As I'm running a bit late and need to get going, I'll leave that for > someone else to respond to in detail -- but there are a lot of extremely > intelligent sysadmins running thousands of machine configurations out > there, many of which who are on these lists. A couple of them have > been working with us to build this. They've mentioned they've wanted > better tools around key exchange, and ways to perform one-offs and for > data gathering -- things that don't fit well into the cfengine/puppet > kind of models. Note that I took exception with the claim that this has anything to do with managing systems, not the tool itself. As I said initially, for casual environments where one-offs are acceptable, func seems fine. But the claims in the announcement go way beyond that. There are plenty of places that try to manage their systems func-style, and when they realize what pain they are inflicting on themselves it is often too late to change course because the cost of switching to a more appropriate tool is prohibitive. > Anyhow, this is part of the reason we're not putting func explicitly on > a systems management email list ... Func has uses other than systems > management, for other apps that want a generic key exchange mechanism > and modular XMLRPC framework. That's what we're building. Why all the claims about 'a better way to manage a large number of systems' and 'fed up with CIM/WBEM' then ? It sounds like 'distributed scripting framework' describes func much better. BTW, a simple, standardized (in Fedora and related distros) way to distribute and manage SSL certs would be really valuable; have you thought about breaking that functionality out separately ? David From berrange at redhat.com Sat Sep 29 01:36:24 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Sat, 29 Sep 2007 02:36:24 +0100 Subject: [et-mgmt-tools] [ANNOUNCE] Func 0.0.12 In-Reply-To: <1191024082.22064.25.camel@localhost.localdomain> References: <46FD6C29.5040809@redhat.com> <1191017778.22064.9.camel@localhost.localdomain> <46FD8A65.3000206@redhat.com> <1191024082.22064.25.camel@localhost.localdomain> Message-ID: <20070929013623.GA27991@redhat.com> On Fri, Sep 28, 2007 at 05:01:22PM -0700, David Lutterkort wrote: > BTW, a simple, standardized (in Fedora and related distros) way to > distribute and manage SSL certs would be really valuable; have you > thought about breaking that functionality out separately ? Such a project recently started in the form of FreeIPA... http://freeipa.org/ There's lots of info on their website, but I guess you could summarize it as an open source ActiveDirectory replacement. 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 lfarkas at bppiac.hu Sat Sep 29 12:13:49 2007 From: lfarkas at bppiac.hu (Farkas Levente) Date: Sat, 29 Sep 2007 14:13:49 +0200 Subject: [et-mgmt-tools] VMGL for centos and fedora Message-ID: <46FE417D.3000608@bppiac.hu> hi, is there any plan to package and/or include VMGL in centos/fedora: http://www.cs.toronto.edu/~andreslc/xen-gl/ -- Levente "Si vis pacem para bellum!" From berrange at redhat.com Sat Sep 29 15:36:07 2007 From: berrange at redhat.com (Daniel P. Berrange) Date: Sat, 29 Sep 2007 16:36:07 +0100 Subject: [et-mgmt-tools] VMGL for centos and fedora In-Reply-To: <46FE417D.3000608@bppiac.hu> References: <46FE417D.3000608@bppiac.hu> Message-ID: <20070929153607.GA19140@redhat.com> On Sat, Sep 29, 2007 at 02:13:49PM +0200, Farkas Levente wrote: > hi, > is there any plan to package and/or include VMGL in centos/fedora: > http://www.cs.toronto.edu/~andreslc/xen-gl/ Fedora is a community driven project - anyone who wishes to package, submit and maintain VMGL is more than welcome to do so...hint...hint... :-) I'm not aware of anyone else doing so for Fedora yet, but there could be someone interested. If you ask again on fedora-devel you might get more feedback. 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 -=|