[DRBD-user] Kernel Oops, RH9.0, 2.4.20-19.9

Tim Hasson tim at aidasystems.com
Sun Jun 13 21:18:04 CEST 2004


I upgraded it yesterday to the latest stable release; 2.4.26
Copied the config from the redhat kernel into the new kernel source, checkout
menuconfig, made dep, bzImage, modules, modules_install, mkinitrd (for scsi and
ext3 modules), and installed.
(2 warnings during depmod -a, related to two modules in kernel/crypto, saw
several people had it - on google - but I guess unmaintained kernel mods?)

After reboot, system booted, all modules working fine, all hardware detected as
it should.

Logged in, cleaned /var/lib/drbd/ directory, recompiled/reinstalled drbd 0.6.12
and started drbd to get a full sync on both drbd0 and drbd1 volumes ( two 35
gig ext3 data partitions )

Sync seemed to go ok for a little while, when I felt some relief, and while I
was keeping an eye on /proc/drbd from drbd1 (pri), I noticed after several
minutes that the drbd1 volume already reached 5% while drbd0 was still at
1.5-2% which is quite unusual since they both usually sync at the same exact
rate..

Then it just froze again. I assume kernel oops.

I asked someone to reboot the server for me since I am not physically there
right now, but will probablly get it done today so I can post the Kernel oops
message / ksymoops trace, if you think it can be of any use..


P.S. I only upgraded the kernel on drbd2, just to try. drbd1 is running the RH
kernel update (older) 2.4.20-19.9
Both drbd1/2 servers are running drbd 0.6.12



Quoting Lars Ellenberg <Lars.Ellenberg at linbit.com>,
on Sun, 13 Jun 2004 11:30:31 +0200:

> / 2004-06-12 17:05:24 -0700
> \ Tim Hasson:
> > Do I have to upgrade the kernel on both drbd1 and drbd2 or can I just
> upgrade
> > the one that is giving kernel oops's (drbd2) ?
> > 
> 
> you don't have to.
> as long as it is the same drbd version,
> it does not care about kernel version or architecture.
> 
> note that changing the kernel may or may not help.
> its only that nobody else has reported an oops related to 0.6.x
> for a long time, many people use it, probably even with your exact
> hardware and kernel. so whatever does cause the oops at your side,
> I seriously doubt that its drbd fault. I even seriously doubt that its
> the RH kernels fault.
> 
> but you said it is easy to reproduce for you (happens every time...).
> so go ahead, and try to change one thing at a time, and see when you can
> no longer reproduce it...  rule out the possible causes one by one.
> 
> > They are both running the same kernel right now (rh 2.4.20-19.9)
> > and also both running drbd 0.6.12
> > 
> > I was hoping I can get away with upgrading the kernel on drbd2 and
> recompiling
> > drbd, because I'd hate to bring down the primary file server for a kernel
> > reboot, and even worse, if the kernel doesn't boot.
> > 
> > SO as of now I am upgrading drbd2 (the server in question) until I get an
> answer
> > from you..
> > 
> > Thank you for all your help..
> _______________________________________________
> drbd-user mailing list
> drbd-user at lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user
> 




More information about the drbd-user mailing list