software RAID1 partitions causing CPU race condition

[T.J. Sullivan]

Hello All,
    I have created 2 RAID1 partitions under YDL 2.2. The creation went
fine and they remount on reboot w/o having to compile a custom kernel.
The problem occurs when I try to use them. I start to rsync or copy
material to them, and it starts off fine. After a short time the machine
stops accessing the drives all the time and appears to access them once
every 5-10 seconds for about 3 seconds. It seems to be causing a 100%
CPU utilization because I can not run commands in other shell either
from telnet or at the terminal. I end up having to shut the machine down
with the power switch. Here are a couple of questions I have.


1. Which logs should I look at for further info? I have gone through the
messages log and perused a number of other ones without success.

2. Is there another disk checking program besides fsck to check the
physical disks? Perhaps an app that can identify bad blocks.

3. Is it the way I set up the md's in the first place? I followed the
directions found here,
http://www.tldp.org/HOWTO/Software-RAID-HOWTO-4.html, and set the block
size during the mke2fs to 4096.

    Any ideas would be helpful. I am now behind schedule for putting
this new server online, and am getting flack that I should just use
windoze/IIS. HELP!!!!!!

Regards,
TJ Sullivan
Systems Administrator
Sonalysts, Inc.