[DRBD-user] BAD! .... messages

Lars Ellenberg lars.ellenberg at linbit.com
Mon Nov 21 18:14:49 CET 2016

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


On Mon, Nov 21, 2016 at 09:31:54AM +0100, Roland Kammerer wrote:
> > >> I found the following in the archives
> > >> (http://lists.linbit.com/pipermail/drbd-user/2012-July/018751.html):

> > >> QUOTE
> > >>   But if you run 8.3 git it is supposed to be fixed, finally...

> > >> UNQUOTE

> > >> I'm running 8.9 ... is this still something I can safely ignore (and not fixed)?
> > > No, you are NOT running drbd 8.9.  That's the version of the userland tools.
> > >
> > > You want to check (and upgrade) your kernel module.
> > > Check: modinfo drbd; head /proc/drbd
> > 
> > Ok. I'm running kernel module 8.4.3 with 8.9 userland tools:
> > 
> > root at ns8p2:~# modinfo drbd
> > filename:      
> > /lib/modules/3.16.0-4-amd64/kernel/drivers/block/drbd/drbd.ko
> > alias:          block-major-147-*
> > license:        GPL
> > version:        8.4.3
> > 
> > ... is this still something I can safely ignore (and not fixed)?
> 
> We are back at what Lars said: "You want to check (and upgrade) your
> kernel module.". Don't miss the upgrade part just because it is in
> parentheses. Out-of-tree 8.4.3 was February *2013*.

Uhm, so the ML post quoted above said "8.3 git" finally fixed "it".

Which is correct. Only that "it" was something different than what
you have with 8.4.3 ...

Even though it leads to similar messages in the kernel log.

BTW, for people finding this in the archives:
we have *different* log lines that prominently contain "BAD",
so just because you see "BAD" does not yet mean you are seeing
the same issue. This one is about
"BAD! sector=12345s enr=6 rs_left=-7 rs_failed=0 count=89 cstate=..."

We've had various issues over the years that would trigger those
log messages, some could safely be ignored, some could lead to
resync not terminating, or other effects.

I'd recommend to not ignore those, but upgrade to latest 8.4.
Yes, these issues are (supposed to be) fixed,
the one I suspect you are hitting was fixed with
 1c378ef drbd: fix resync finished detection
Which was somewhere between 8.4.4 and 8.4.5 in 2014, and yes, it was a
different issue from the one referred to in the ML post you found.


-- 
: Lars Ellenberg
: LINBIT | Keeping the Digital World Running
: DRBD -- Heartbeat -- Corosync -- Pacemaker

DRBD® and LINBIT® are registered trademarks of LINBIT
__
please don't Cc me, but send to list -- I'm subscribed



More information about the drbd-user mailing list