[Linux-cluster] Running GFS without fencing and maybe locking ; -)
David Teigland
teigland at redhat.com
Thu Mar 23 18:19:02 UTC 2006
On Thu, Mar 23, 2006 at 10:42:00AM +0100, Arnd wrote:
> So even if I'm adding two hosts to the cluster with write access then
> how can I control, that if one of these nodes fails, the other node with
> write access is replaying the journal (and not the node which is doing
> the fencing and has the filesystem mounted readonly)?
This looks like another problem in gfs. We need can't have a readonly
node potentially preventing a rw node from doing recovery. (The ro node
shouldn't be taking the journal lock if it can't eventually do the
recovery.)
> Ok, back to the tests. Both nodes (adnux2 and adnux3) weren't able to
> access the filesystem after adnux4 failed:
>
> adnux3 data # ls -l /home/data/gfs
> ls: /home/data/gfs: Input/output error
This looks like the fs might have been withdrawn (shutdown) after gfs
found some inconsistency.
Dave
More information about the Linux-cluster
mailing list