Am 24.08.2011 16:17, schrieb Lars Ellenberg:
Same happens with drbd83 and drbd84 packages from elrepo and with a
self-compiled drbd84 from linbit sources.
I don't think this is anything DRBD specific.
The Oops happens from drbdadm, which is a normal user space tool,
when doing some ioctl on apparently some socket.
Doing some ioctl on some socket file descriptor from userland
should not be able to trigger an oops.
Try to search for similar symptoms not involving DRBD.
I hope I sorted it out by re-creating the cluster with cman plus
corosync and pacemaker and without the dlm-pcmk and gfs-pcmk packages.
The drbd config however remained the same. Whenever I reboot single
servers now or go into standby, no further kernel panics occur.
[...]
This is a "fair weather setup". It will fail (aka: behave in strange
and unexpected ways) when things go wrong.
Getting a DRBD dual-primary cluster file system setup to work reliably
in face of errors is a bit more complex.
And you really need fencing (stonith).
You are right, I forgot to mention that this is a test setup.
Regards,
Peter
_______________________________________________
drbd-user mailing list
[email protected]
http://lists.linbit.com/mailman/listinfo/drbd-user
