koan 2.2.3 and fedora17 with grub2
by Ubay Dorta Guerra
Hello,
I have a fedora17 client with koan 2.2.3 and i want to reinstall the
system with te following command:
[root@client1 ~]# koan --server server1 --system=client1 --replace-self
...
- ['/sbin/grubby', '--add-kernel', '/boot/vmlinuz_koan', '--initrd',
'/boot/initrd.img_koan', '--args',
'"ks=http://server1/cblr/svc/op/ks/system/client1 ksdevice=link
kssendmac lang= text "', '--copy-default', '--make-default',
'--title=kick1347011477']
...
- reboot to apply changes
When the machine reboots, it is starting the previous system and not
the installation process.
The koan command uses grubby and fedora17 now uses grub2 as default.
Looking at: /usr/lib/python2.7/site-packages/koan/app.py:
1074
1075 # Set paths for Ubuntu/Debian
1076 # TODO: Add support for other distros when they
ship grub2
1077 if make in ['ubuntu', 'debian']:
1078 grub_file = "/etc/grub.d/42_koan"
1079 grub_default_file = "/etc/default/grub"
1080 cmd = ["update-grub"]
1081 default_cmd = ['sed', '-i',
's/^GRUB_DEFAULT\=.*$/GRUB_DEFAULT="%s"/g' % name, grub_default_file]
Is it going to be supported fedora with grub2 for the next version
of koan?
Meanwhile i have changed manually grub2 like the script for
debian/ubuntu.
Thanks in advance.
Regards.
11 years, 8 months
Re: [cobbler] xenserver 6.0.2 is stuck while pxe booting
by Srikanteswararao Talluri
On Mon, Sep 10, 2012 at 3:08 AM, Srikanteswararao Talluri <
kanti.sri(a)gmail.com> wrote:
>
>
> On Mon, Sep 10, 2012 at 3:08 AM, Srikanteswararao Talluri <
> kanti.sri(a)gmail.com> wrote:
>
>> Thanks, it is working now
>>
>>
>> On Mon, Sep 10, 2012 at 3:06 AM, Srikanteswararao Talluri <
>> srikanteswararao.talluri(a)citrix.com> wrote:
>>
>>> I found that Problem is with mboot.c32, If I use mboot.c32 file from xen
>>> 56 source, I could successfully pxe boot xen 6.0.2. For some reason
>>> mboot.c32 provided with xen6.0.2 is not working****
>>>
>>> ** **
>>>
>>> *SrikanteswaraRao Talluri*
>>>
>>> ** **
>>>
>>
>>
>
11 years, 8 months
Cobbler DNS management
by Bai Shen
I have cobbler set to manage the dhcp and dns. The dhcp server handing out
the cobbler servers ip, but I'm not able to resolve anything.
Should the dhcp be set to us the regular dns servers or am I doing
something wrong on the cobbler side?
Thanks.
11 years, 8 months
Koan + SL 6.2 no ks file
by Ronald J. Yacketta
Hello all!!
Just installed a new HP Blade and put SL 6.2 on it with the latest koan
(2.2.3) and noticing that koan is not tacking on --extra-args when
attempting to generate an new vm:
koan --virt --server=da.server --profile=SL-6.2-x86_64
--virt-name=tester --virt-path=/mnt/sas3_1/ -w &
[1] 870
[bs3-1 ~]# - looking for Cobbler at http://eden.man:80/cobbler_api
- reading URL: http://10.137.31.1/cblr/svc/op/ks/profile/SL-6.2-x86_64
install_tree: http://10.137.31.128/scientific6-64
libvirtd (pid 14631) is running...
- adding disk: /mnt/sas3_1//tester-disk0 of size 25 (driver type=raw)
- ['virt-install', '--connect', 'qemu:///system', '--name', 'tester',
'--ram', '512', '--vcpus', '1', '--autostart', '--vnc', '--virt-type',
'qemu', '--hvm', '--location', 'http://10.137.31.128/scientific6-64/',
'--arch', 'x86_64', '--os-variant', 'generic26', '--disk',
'path=/mnt/sas3_1//tester-disk0,size=25,driver_type=raw', '--network',
'bridge=br9', '--wait', '0', '--noautoconsole']
Did a quick search and found the following bug from 2009
w(https://bugzilla.redhat.com/show_bug.cgi?id=853980) hich describes
exactly what I am seeing.
Anyone else seeing this? better yet, has this been reported (other then
back in 2009)?
-Ron
11 years, 8 months
Fedora 17 DVD cobbler import fail
by Ubay Dorta Guerra
Hello,
I have installed a Fedora 17 x86_64 server with cobbler-2.2.3. I am
following the manual and i'm trying to import the DVD:
[root@server1 ~]# mount -t iso9660 -o loop,ro
/tmp/Fedora-17-x86_64-DVD.iso /mnt/dvd/
[root@server1 ~]# cobbler import --name=fedora17 --arch=x86_64
--path=/mnt/dvd/
task started: 2012-09-06_100837_import
task started (id=Media import, time=Thu Sep 6 10:08:37 2012)
!!! TASK FAILED !!!
The log file /var/log/cobbler/cobbler.log:
Thu Sep 6 10:08:37 2012 - DEBUG | REMOTE CLI Authorized; user(?)
Thu Sep 6 10:08:37 2012 - INFO | REMOTE start_task(Media import);
event_id(2012-09-06_100837_import); user(?)
Thu Sep 6 10:08:38 2012 - INFO | import_tree; ['/mnt/dvd/', 'fedora17',
None, None, None]
Thu Sep 6 10:08:38 2012 - INFO | importing from a network location,
running rsync to fetch the files first
Thu Sep 6 10:08:38 2012 - INFO | running: rsync -a '/mnt/dvd/'
/var/www/cobbler/ks_mirror/fedora17-x86_64
--exclude-from=/etc/cobbler/rsync.exclude --progress
Thu Sep 6 10:08:38 2012 - INFO | received on stdout: sending
incremental file list
./
sent 29 bytes received 15 bytes 88.00 bytes/sec
total size is 0 speedup is 0.00
Thu Sep 6 10:08:38 2012 - DEBUG | received on stderr:
Thu Sep 6 10:08:38 2012 - INFO | No import managers found a valid
signature at the location specified
Is there something wrong when i use "cobbler import"?
The rsync command in log file works:
[root@server1 ~]# rsync -a '/mnt/dvd/'
/var/www/cobbler/ks_mirror/fedora17-x86_64
--exclude-from=/etc/cobbler/rsync.exclude --progress
sending incremental file list
created directory /var/www/cobbler/ks_mirror/fedora17-x86_64
./
.discinfo
35 100% 0.00kB/s 0:00:00 (xfer#1, to-check=1226/1228)
.treeinfo
925 100% 903.32kB/s 0:00:00 (xfer#2, to-check=1225/1228)
TRANS.TBL
1753 100% 855.96kB/s 0:00:00 (xfer#3, to-check=1224/1228)
EFI/
EFI/TRANS.TBL
.....
Thanks in advance.
Regards.
11 years, 8 months
Can Koan be used to pxeboot a vCenter 5 VM?
by Daniel Kertby
Hi all.
I have read the man manual page for Koan and also searched for an
answer but haven't been wiser.
Is it possible to use koan to reset (or power on if off) then issue a
pxeboot for a vCenter 5 VM?
Im not interested to let Koan create the VM.
Thanks in advance,
Daniel
11 years, 8 months
tftpd.py not working after upgrading to 2.2.3
by Lassi Pölönen
I recently upgraded my cobbler server from rhel6.2 to 6.3 and apparently
cobbler had updated as well from cobbler-2.2.2-1.el6.noarch to
cobbler-2.2.3-2.el6.noarch. The problem is that clients now seem to get
tftp timeout.
Looking at the abrt generated backtrace:
templar.py:59:__init__:AttributeError: 'NoneType' object has no
attribute 'api'
Traceback (most recent call last):
File "/usr/sbin/tftpd.py", line 1147, in <module>
sys.exit(main())
File "/usr/sbin/tftpd.py", line 1125, in main
templar = cobbler.templar.Templar(None)
File "/usr/lib/python2.6/site-packages/cobbler/templar.py", line 59,
in __init__
self.api = config.api
AttributeError: 'NoneType' object has no attribute 'api'
Local variables in innermost frame:
logger: None
config: None
self: <cobbler.templar.Templar instance at 0x143ef80>
A bug in here or am I missing something? It used to work fine before and
if I change tftp module to in.tftpd, the clients will work. That's just
not a feasible option since it takes awfully long with the amount of
systems..
11 years, 8 months
koan not injecting a kickstart file when creating virtual machine
by Tim Coote
Hullo
I've recently preupgrade-cli'd a box from f15 to f17 and now my koan/cobbler builds no longer inject kickstart files.
Even if I take the stock /var/lib/cobbler/kickstarts/sample.ks and use it as the ks for a profile, when I run:
sudo koan --virt --virt-name=robot --server=puppet --system=robot
[robot's defined as being based on f17-i386 profile, which has a Kickstart using the sample.ks above]
After powering up, the vm starts a graphical install and seems to have no knowledge of the kickstart file. If I take it through the graphical install, the /root/anaconda-ks.cfg is the default created by anaconda. Nothing gives the game away when I log into the VM's console during the installation.
It doesn't seem to matter if I use an older profile (I've tried F15).
I'm using cobbler-2.2.3-2.fc17.noarch, and koan-2.2.3-2.fc17.noarch
sudo cobbler validateks doesn't complain (I had to fix the $kickstart_start/$kickstart_done in the old kickstarts, but these are not present in the sample.ks).
The only oddities that I can identify are /var/log/messages entries like these:
Sep 3 10:28:18 mercury libvirtd[4807]: 2012-09-03 09:28:18.284+0000: 4808: error : virSecurityDACRestoreSecurityFileLabel:148 : cannot resolve symlink /var/lib/libvirt/boot/virtinst-vmlinuz.
5LB4JW: No such file or directory
Sep 3 10:28:18 mercury libvirtd[4807]: 2012-09-03 09:28:18.284+0000: 4808: error : virSecurityDACRestoreSecurityFileLabel:148 : cannot resolve symlink /var/lib/libvirt/boot/virtinst-initrd.i
mg.fhjuOM: No such file or directory
which could imply that there's something odd that's preventing the libvirtd setting up the boot filesystem, but I don't know enough about the innards to work that out. sudo rpm -V libvirtd shows no errors.
Any pointers on how to work out what's going wrong/how to identify what's going wrong?
Tim
11 years, 8 months