99-raid-check

[BruceLee]

sure:
# uname -a
Linux servername 2.6.32.8-1.art.x86_64 #1 SMP Sat Feb 13 19:13:41 EST 2010 x86_64 x86_64 x86_64 GNU/Linux

[premierhosting]

Troy, what's the output of your uname -a ?

BruceLee, we seem to have some minor version differences in the Kernel. Seems like a place to look.

[mikeshinn]

Its not the kernel. Its what that script does:

http://lists.centos.org/pipermail/cento ... 86653.html

[Troy McClure]

Yes, it is nothing to do with the kernel. I have used different ones. But we have different issues I think. It never completes when I run mine. It locks up. I manually ran a repair on the raid array last night and it completed successfully. I plan to run a check later tonight to see what happens when the check runs now.

[Troy McClure]

Well, funny enough, I ran a check on the raid array last night and it completed successfully. I didn't run it using the 99-raid-check, I just manually started the check. Then today around 13 hours after it completed, my server hung. I had to reboot it to get it back online.

[mikeshinn]

Did it hang, or was load just thru the roof? Checking the RAID will naturally drive i/o up for all the raids (and channels) being checked, so is it possible your i/o bus was just hammered?

[Troy McClure]

Well, this was 13 hours after the check completed, so I assume the load was normal. Unfortunately I was not monitoring the server when it happened. I just found out when one of my clients called about it being down. I was having the problem before when the check was running, but now it seems to be acting like premierhosting's server where it just hangs after 12 to 13 hours of running the check.