Red Hat Bugzilla – Bug 57906
Bad: System crashes with gdth raid controller
Last modified: 2007-04-18 12:38:51 EDT
From Bugzilla Helper: User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; Q312461) Description of problem: I don't have much details. I installed a new server. I upgraded to 2.4.9- 13enterprise kernel using up2date. After a day it first hung. Even the switchbox it was connected to sort of "hung" it that server was selected. Today, same thing. Ping is still working although. Gonna take a look tomorrow. Dual PIII system with gdth raid controller. 1 GB RAM. Very annoying. Version-Release number of selected component (if applicable): How reproducible: Always Steps to Reproduce: 1. Turn it on 2. Way a day or two 3. Push the reset button Additional info:
It sounds odd and it may be a coincidence but I was connected via SSH with the F-Secure Windows SSH Client. My laptop unexpectedly went into hibernation while the connection etc. was still up. I immediately rebooted and then I couldn't reach the server anymore. I can ping it but nothing else.
Need more info? Take a look at RHN profile cents.iip.lu Sorry, but I'm anxious about getting this fixed!! Happy new year BTW
Can you try adding "mem=800M" on the kernel commandline (in /boot/grub/grub.conf on the vmlinuz line, if you use grub)
Also, it might be interesting to try the kernel at http://people.redhat.com/arjanv/testkernels which has a few bugs fixed; one could be hit on your machine....
Thanks, I have more information. The console was still there. One could enter the username but not anymore the password. All the consoles were full of this errors: ext3-fs error device 08:03 not ready I/O error: dev 08:03 sector 0
I found something that might be related http://groups.google.com/groups? q=kernel+gdth++bug+2.4&hl=en&scoring=d&rnum=1&selm=linux.smp.OE26fLJGWSa1aNEVJpj 0000805b%40hotmail.com I can't *try* much since this is a production system. Is the gdth driver up to date? The latest version is 2.03 which is revision 1.57
I'm afraid to say the new test kernel didn't help. I have no debugging info either. The screen was black when it crashed 30 mins ago. I think I'm gonna try the normal (not SMP) kernel now.
I downgraded to kernel-2.4.9-13.i586.rpm. With append "mem=800M". For now its stable. Note that 2.4.9-17enterprise also crashed. I'll keep you advised.
The standard i586 kernel is still running fine...
2.4.9-21 (NOT SMP) without 800M setting also running fine must be SMP related (didn't try the new SMP one though)
Seems to be OK using 2.4.9-31smp