What the I think the kernel update broke on Open suse11.1

All,

So far I’ve had two major breakages due to the vmlinuz-2.6.27.19-3.2-pae kernel update on my Opensuse 11.1 web server.

I’m sharing my problems and fixes for 2 reasons.

1.There is a lot I don’t understand and maybe it is not the kernel update that broke these, if not then I appreciate any enlightenment offered.

2.If it is the kernel update then I want other people to know so they don’t struggle for hours on end as I have.

The first thing that I noticed was that upon boot the screen asked me if I wanted to fall back to the /dev/disk/by-id/scsi-SDELL_Drive#0_D3E2F33B-part3. I had to answer Y before the boot continued.

The previous kernel’s menu.list had the string “root=/dev/disk/by-id/scsi-SDELL_Drive#0_D3E2F33B-part3” on the kernel line. The new version truncated that string to “root=/dev/disk/by-id/scsi-SDELL_Drive” a drive that does not exist.

(I’ve lost the actual boot.msg file that contains this since I overlaid it as I rebooted numerous time to troubleshoot and test.)

My fix was to change the root= parameter on the kernel line of the /boot/grub/menu.lst to read root=/dev/sda3.

I tried changing it to the previous root=/dev/disk/by-id/scsi-SDELL_Drive#0_D3E2F33B-part3 but then the boot would hang and I get a maintenance shell.

My second problem was that my web server would no longer check out a dhcpd address. After much futzing I did two things.

1.I switched networking methods from NetworkManager to Traditional using curses yast2->Network Devices->Network Settings.

2.When that didn’t help I used curses yast2->Network Devices->Network Settings to statically assign an ip address.

At this point in time I need to leave the web server in production to provide web pages to our employees. If all is well today I’ll reboot it tonight to make sure it will survive a reboot.

If I discover more problems and solutions I’ll share them.

Sincerely,

Park Richard

I think I caused the breakage. I have cron use zypper to update starting at 4:22 am on Wednesday and then I have a reboot at 5 am. Can’t be a good scenario. I’m looking into how to run the installs again.

Hum, that sounds dangerous. What happens if the reboot starts before the update is complete? I suspect a corrupt system.

Really, the only time you really need to reboot is after a kernel update. For other updates, you may need to restart some services.

If you want to reboot after an update, do the reboot in the same script that soes the update.