Simpler Navigation for Servers and Operating Systems - Please Update Your Bookmarks
Completed: a much simpler Servers and Operating Systems section of the Community. We combined many of the older boards, so you won't have to click through so many levels to get at the information you need. Check the consolidated boards here as many sub-forums are now single boards.
If you have bookmarked forums or discussion boards in Servers and Operating Systems, we suggest you check and update them as needed.
cancel
Showing results for 
Search instead for 
Did you mean: 

Kernel Crash- DL380G3-linux 7.3

SOLVED
Go to solution
ryan_143
Occasional Visitor

Kernel Crash- DL380G3-linux 7.3

Hi,

I have a server that goes down twice a week from a server panic in Linux 7.3 kernel 2.4.18-3smp. I've attached an image of the crash. Can anyone tell me what I could to do to correct this? Appreciate any help in this matter...

Ryan
5 REPLIES
Alexander Chuzhoy
Honored Contributor

Re: Kernel Crash- DL380G3-linux 7.3

I'd try to install a newer kernel and see if this problem persists.
Try to download the RPM of newer smp kernel
and to install it with rpm -ivh kernelversion

Best regards...
Mark Grant
Honored Contributor

Re: Kernel Crash- DL380G3-linux 7.3

This seems to be an error in the kernel filesystem journaling routines. If the error really is in commit.c as is suggested, then this is in the code to commit to the log in the ext3 filesystem. You may have a marginally corrupt fileystem.

This obviously shouldn't happen. If you need a quick fix, either upgrade, re-create your ext3 filesystem or try a different fileystem type e.g reiserfs.

However, you should really download a vanilla Linus Torvalds kernel and try and re-produce the error. If you can, you should submit a bug report to the linux kernel mailing list or direct to Stephen Tweedie.
Never preceed any demonstration with anything more predictive than "watch this"
Steven E. Protter
Exalted Contributor
Solution

Re: Kernel Crash- DL380G3-linux 7.3

2.4.20 is a little more stable, fixes some bugs.

Upgrading to that has improved my reliability on my 7.3 box.

SEP
Steven E Protter
Owner of ISN Corporation
http://isnamerica.com
http://hpuxconsulting.com
Sponsor: http://hpux.ws
Twitter: http://twitter.com/hpuxlinux
Founder http://newdatacloud.com
ryan_143
Occasional Visitor

Re: Kernel Crash- DL380G3-linux 7.3

Do appreciate all the feedback... and we were going to upgrade the kernel initially but the kernel version that hp drivers support is the kernel that I have installed under 7.3 and hp support has verified this. The problem is that if by upgrading the kernel to a more stable one, i might run or inherit more problems due to that hp has not tested their drivers under a newer kernel. Is it safe to upgrade to a kernel that hp drivers have not been tested on? Are your experiences more positive than negative in this matter? Thanks again for all your answers...

Ryan
Mark Grant
Honored Contributor

Re: Kernel Crash- DL380G3-linux 7.3

Ryan, Generally speaking, upgrading a single point release, for example from 2.4.18 to 2.4.19 will have no significant issues for user space applications. This is why most application vendors say they will support Linux 2.2 or 2.4 as the interfaces won't change significantly. However, for drivers, the situation may be completely different. If they are open source drivers, or you at least have the source code, you would re-compile them after your upgrade and they will almost certainly work. Binary drivers probably won't work though, they might, but probably won't. If you have a driver that will recompile when built then the upgrade is a good thing for you to try. It's probably this driver that is causing you the trouble in the first place. If you have a binary only driver then you need to tread carefully and push HP to supply a driver for a more recent version of the kernel. In the meantime, see if you can find any patterns in your machine failures and perhaps re-create your ext3 filesystems or, as I mentioned above, change to reiserfs (it's faster on SMP systems anyway)
Never preceed any demonstration with anything more predictive than "watch this"