[oclug] How do you read Ext2-fs errors?

[Dean Staff]

Hi all, I'm hoping someone the list can help me solve this problem.

I have a PIII600 Dell box with PERC 3 SCSI Raid controller. It has 2 
drives that are setup as raid 1. I'm running Redhat 6.2. 

All had been running well, until late last week. when I started 
getting the following console errors messages.

EXT2-fs error (device sd(8,5)): ext2_readdir: directory #308562 
contains a hole at offset 1783365632

I get lots of these messages, but the only thing that changes is the 
offset #.

This all started just shortly after we added a new cron job that 
grep's through the popper.log to make sure that other servers are 
popping regularly. the script only checks that tail of the log file, 
but the log file is pushing 540Mb in size. 

If I shutdown and reboot once I get these errors I'm forced to run 
fsck. Durring that I promted to fix dtime (?) errors. Everything goes 
smoothly and I reboot the system with no problems. 2 days later...I 
start all over. 

I've used the disk utilities on the RAID Controller to verify disk 
media, and I get no errors I've even run e2fsck several time in a row 
with the -f switch, and there are no subsiquent errors found.
Is is possible that the size of the log file is causing the error?
Or is my one of the HDD's ready to crash?

Thanks in advance
Dean
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Dean Staff
Protus IP Solutions
210 - 2435 Holly Lane
Ottawa, ON K1V 7P2 Canada
613-733-0000 ex 546 Fax 613-248-4553
e-mail: dean at protus.com Web: http://www.protus.com
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~