[DRBD-user] SLES9 .244 drbd 8.2.0 kernel oops

Florian Haas florian.haas at linbit.com
Mon Oct 29 15:56:50 CET 2007

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


On Monday 29 October 2007 15:30:49 Andre Konopka wrote:
> Florian Haas wrote:
> > Andre,
> >
> > just making sure: you changed your drbd.conf on both servers, rebooted
> > both servers, and _after_ both servers came back up, you attempted
> > another disconnect and got a kernel oops from that?
> >
> > If yes, please post the relevant lines from your most recent kernel log
> > again (from both servers).
>
> I only tried it on the first server....on the second one drbd was unloaded.
> I did the same test with 0.8.6 (only one node aktiv). No trouble
> at all.

Yes, we know this is an 8.2.0 only issue. It doesn't apply to 8.0.6.

> Oct 29 14:39:43 pdxrs001 kernel: floppy0: no floppy controllers found
> Oct 29 14:40:57 pdxrs001 kernel: drbd0: Discarding network configuration.
> Oct 29 14:40:57 pdxrs001 kernel: drbd0: tl_clear()

Wait a minute. For your drbd0, contrary to your earlier post, discarding the 
network config is in fact now working quite nicely.

> Oct 29 14:40:57 pdxrs001 kernel: drbd1: Discarding network configuration.
> Oct 29 14:40:57 pdxrs001 kernel: drbd1: tl_clear()
> Oct 29 14:40:57 pdxrs001 kernel: Unable to handle kernel NULL pointer
> dereference at 0000000000000000 RIP:
> [...]

Now it's your drbd1 that's causing the oops. You didn't post your config for 
your drbd1, so what's up with that? If your drbd1 doesn't have a 
cram-hmac-alg and shared-secret, you'll run straight into the same 8.2.0 
bug. :-)

Hint: not a bad idea to put that cram-hmac-alg and shared-secret into 
your "common/net" section.

Cheers,
Florian

-- 
: Florian G. Haas   
: LINBIT Information Technologies GmbH
: Vivenotgasse 48, A-1120 Vienna, Austria



More information about the drbd-user mailing list