[Linux-cluster] GFS volume already mounted or /mnt busy?

[Robert Peterson]

bigendian+gfs at gmail.com wrote:
> Thanks Robert.  I'm sorry, but I put you a little off track do to a 
> typo.  I meant to say that gfs_fsck completes without errors.  I've 
> run it twice.  The first pass did make a number of corrections, but 
> completed through stage 5.  A subsequent gfs_fsck completed with no 
> errors.
>
> The "gfs_tool sb /dev/etherd/e1.1 proto" does return lock_dlm as the 
> lock protocol.
>
> Which package contains the gfs_edit tool?
>
> Thanks again,
> Tom
Hi Tom,

I'm not directly involved in the packaging, so I don't know what package 
gfs_edit is in.
It might not even be packaged at all, but only in our cvs source tree.

Ordinarily we don't do this, but why don't you send me the first 1M of your
file system and I'll see if I can tell whether it's been overwritten 
somehow.  So:
dd if=/dev/etherd/e1.1 of=/tmp/bob bs=1M count=1
and email /tmp/bob to me.

Regards,

Bob Peterson
Red Hat Cluster Suite