[Linux-cluster] failover questions after upgrade

[Lon Hohberger]

On Mon, 2006-11-27 at 20:43 -0500, jason at monsterjam.org wrote:
> so the main mountpoint for my gfs filesystem is /mnt/gfs, but my apache htdocs root is at 
> /mnt/gfs/htdocs
> so after I changed my /etc/cluster/cluster.conf 
> to be 
> <fs device="/dev/mapper/diskarray-lv1" fstype="ext3" mountpoint="/mnt/gfs" name="apache_content"/>
> 
> my /mnt/gfs does not get mounted anymore when the cluster servers start up.. whats the correct way 
> to do this?

It looked like it was getting mounted elsewhere, like in fstab.  The
logs indicated thusly:

- <warning> Device /dev/mapper/diskarray-lv1 is mounted 
on /mnt/gfs instead of /mnt/gfs/htdocs 

That meant, I think, that you had it configured to be mounted
on /mnt/gfs/htdocs, but it was currently mounted on /mnt/gfs.  The
cluster got confused and tried to unmount it, and that didn't work.

I kind of missed the fact that the file system was ext3 the first time
around; in that case, you should *not* have it in /etc/fstab at all --
otherwise it could be mounted on both nodes on startup (very bad - and
it could possibly explain why unmounting it failed...!).

-- Lon
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 189 bytes
Desc: This is a digitally signed message part
URL: <http://listman.redhat.com/archives/linux-cluster/attachments/20061204/0f4f08fb/attachment.sig>