[Linux-cluster] no version for "gfs2_unmount_lockproto"

[Ferenc Wagner]

Christine Caulfield <ccaulfie at redhat.com> writes:

> Ferenc Wagner wrote:
>> "Marc - A. Dahlhaus [ Administration | Westermann GmbH ]" <mad at wol.de>
>> writes:
>> 
>>> Hello Ferenc,
>>>
>>> Am Freitag, den 15.02.2008, 03:56 +0100 schrieb Ferenc Wagner:
>>>
>>>> How could I find out about a pending fence?  Only by periodic messages
>>>> in the syslog?
>>> "cman_tool nodes" is your friend here.
>>> Look for a status value of "F".
>>> The man page of cman_tool has more info on this.
>> 
>> Thanks!  Both live nodes have status M (member?) now, so this doesn't
>> seem to be the problem.  My man page doesn't describe the meanings of
>> the flags, unfortunately...
>
> Later versions of cman_tool have the flags explained in the man page.
> There is no 'F' flag, but you can get some fence details from the command
>
> cman_tool nodes -f

Ah.  For me, cman_tool nodes produces the exact same output as
cman_tool nodes -f.  This is with cluster-2.01.00.  But thanks for the
useful info.

Anyway, problem solved: it was binaries (mainly mount.gfs) ending up
in /usr/sbin instead of /sbin.  After correcting that, all three nodes
are up and happily sharing their GFS.  Thanks for the help, everybody!
I certainly learnt a lot during troubleshooting this.
-- 
Regards,
Feri.