Hi all,
I'm seeing frequent oopses when using a combination of RAID1 plus
ext3fs.
Here is my machine config:
PIV 1.5GHz, UP, 256MB RAM, 4 RAID1 devices (md0=/ 1GB, md1=/var 5GB,
md2=/usr 7GB, md3=/home 60GB), 2 80GB IDE disks (on separate
controllers).
My stress test consists of running:
while true; do make clean; make -j10 bzImage modules; done
on a linux-2.4.xx source tree placed in the /home partition. The oopses
happen within a few hours of starting the test.
I have also tried downgrading to linux-2.4.22 but I got the same
results. Currently I've switched to using ext2 on all partions but I
haven't tested it enough to reach a conclusion.
Attached are the kernel .config file plus /var/log/messages (with boot
messages and oops).
Is this a known problem? Any advices to have a stable setup using RAID1
config?
TIA,
Rui
Hi,
On Wed, 2004-04-21 at 15:05, Rui Sousa wrote:
> I'm seeing frequent oopses when using a combination of RAID1 plus
> ext3fs.
The log is full of errors like
EIP: 0010:[prune_dcache+24/320] Not tainted
Unfortunately, dcache corruptions like that are most often a sign of bad
hardware, usually memory. I'd recommend an overnight memtest86 run as
the first step towards diagnosing the problem.
Cheers,
Stephen