OVH Community, your new community space.

updating proxmox


zimsters
18-10-2009, 16:54
ok the exact words were "Software config rescue'

have updated my ticket asking what they did, will report back

Myatu
18-10-2009, 12:11
"software fix" is quite descriptive indeed! Would be nice to know what they actually do for future reference (would save them some time too).

zimsters
18-10-2009, 12:07
ovh fixed it, no idea how. it raised an incident, which got dealt with in about 5 minutes with a reply "software fix" or something liek that, and she's back online...

Myatu
18-10-2009, 11:54
Not good indeed. I'd revert back to the previous kernel for now. Seems to be a plethora of possibilities on this one.

zimsters
17-10-2009, 12:26
any idea how to fix this issue? this is a server udpated using above method. this is what i see via vkvm when i boot via harddisk.



if i try on netboot with root=/dev/sda1 there's a kernel panic and she doesn't boot.

ideas?

freshwire
11-10-2009, 21:45
Yeah the network kernel can be booted or you can use rescue mode and fix the problem.

gigabit
11-10-2009, 20:10
I'm not sure, but I think if there server doesn't come back up, you would be able to use one of the network kernels to boot instead - on the page with rescue mode there are some kernels in the list. It has helped me before when a server didn't like its kernel so i was able to get it back online and fix the main kernel and then change back to boot from hard drive.

zimsters
11-10-2009, 19:49
guys i really need to reboot this server but i don't know if it'll come back up. can you please tell me what i need to do to ensure that the server comes back up with the proper kernel? and what to do if it doesn't and i'm in the rescue mode?

any help much appreciated.

zimsters
11-10-2009, 12:52
"double check the date against that of the kernel" - meaning the modified date?

when I do an ls -l of the directory in question, I'm comparing:

- initrd.img-2.6.24-8-pve (modified 10-10-2009 12:44)
-vmlinuz-2.6.24-8-pve (modified 06-10-2009 11:50)


re: update-initramfs, is this server at least set on this front? so right now it's just ensuring the kernel is ok for reboot?

Myatu
10-10-2009, 21:39
Seems right, yes.

The file I mentioned is the /boot/initrd.img-$(uname -r) file; double check the date against that of the kernel /boot/vmlinuz-$(uname -r).

The fact that update-initramfs actually showed mdadm is good, because this is not included with the Proxmox distro (but *is* with OVH for Softraid, and very much needed indeed).

As for fixing the boot issue, that's reverting to the previous kernel. You could adjust the /boot/menu.lst to include the older kernel files, which can make things a lot easier. But Proxmox had named the kernel the same as the previous one in 1.3 (for the security update), which they shouldn't have done as it overwrote the old kernel files. So just to be sure, make a manual backup copy first.

zimsters
10-10-2009, 12:25
ok so as to turn this into an "upgrade proxmox safely" guide, is the following accurate:

1) edit /etc/apt/sources.list
pico /etc/apt/sources.list
2) Add
deb http://ftp.debian.org/debian lenny main contrib

# PVE packages provided by proxmox.com
deb http://download.proxmox.com/debian lenny pvetest

# security updates
deb http://security.debian.org/ lenny/updates main contrib
3) run
apt-get update

apt-get dist-upgrade

4) run update-initramfs this is for 1.4b2
update-initramfs -tuk 2.6.24-8-pve

Questions:
- is the initramfs command written above accurate? getting following output:
update-initramfs: Generating /boot/initrd.img-2.6.24-8-pve
W: mdadm: /etc/mdadm/mdadm.conf defines no arrays.
W: mdadm: no arrays defined in configuration file.
- how to check the file before rebooting? what file?
- how to fix this issue in case you reboot and it fails. i.e. procedure for rescue mode once you're in rescue.


May as well turn this into a definitive guide

Myatu
10-10-2009, 09:00
Not sure about the beta, but the issue's been the ramdisk image. So make sure you use "update-initramfs" and double check the file before you reboot, otherwise you may end up with an unbootable system (thank goodness for Rescue Mode).

Just make sure you make a full backup of your containers (with vzdump) and important settings before you do this. Standard CYA procedure

zimsters
10-10-2009, 04:40
hey guys,

i know there's some "special" way to upgrade proxmox rather than the very simple steps given by proxmox, thanks to ovh's "unique" way of handling kernels etc.

Anyone have a step by step of what's needed? Thanks!

ideally want to get 1.4b2.