5

openSuSe 11.2 update failed and now server won't boot - urgent help needed!

view full story
linux-howto

http://forums.opensuse.org – Hi, I really hope someone can give me help with the problem I have with my server. I had problems with dual screen setup which I couldn't fix properly and found that in 11.2 there is xrandr tool used instead of SaX2. Problem started when I updated from 11.1 to 11.2 using network update (downloaded ISO image without rpms - 11.2 has been released some time ago so most of the rpms would had to be updated anyway thus why not from the very beginning.. major mistake, it seems). Update started fine. Something like 1300+ rpms were being updated. I let the computer do the thing (couple of hours). When I eventually returned to check how is the update going I found out that computer had crashed. System has been very solid (except ATI's graphic driver part) thus I highly doubt any HW problems. After reset tried to repair installation and some things got fixed but not all (noscp and pattern_openSUSE_enhanced.. were never installed - maybe something else as well). Tried rebooting: blank. Retry on repair, boot: blank (despite whether openSUSE or openSUSE failsafe was chosen). Then tried installing again. Installed several missing rpms (oss and non-oss). Reboot: blank. After several automatic and manual fixes (which always left those couple of rpms uninstalled even though it said: downloading (100%), installing (100%), updating files) the result remained the same: blank. Then I managed to get another computer connected to the network, downloaded full 11.2 image and tried to repair from DVD. No luck. Same situation. Repair tool also pretty much always find errors from sda3 (system disk) and fixes them. And always finds errors from bootloader, fixes them, finds same errors again and fixes them again. Are these error real or not, if so are they fixed or not? Last tests what I did was BIOS setting changes. PnP OS was set to disable. ACPI support disabled. Also changed failsafe kernel parameters (noacpi and vga=textmode). After that I was able to see that kernel is indeeded loaded and started to some extent. Start up ends in: [0.30xxxxx] io scheduler cfq registered So, can anyone help with the problem? Most of the important data is luckily on RAID arrays (sdb and sdc) but unfortunately /etc/* is on sda3. I managed to take tar-backup of it but it still resist on same drive (several copies: /tmp, /opt, /home.bak). It's important to get the server back up online because my e.g. my mail server is down and cannot receive any mails at the moment.. Mboard: Asus M3A-H/HDMI CPU: AMD 4850X2 (Distributions)