Greetings list,

I have been trying, quite unsuccessfully, to setup a HA NFS server
using SLES 10. I have looked at various articles in the linux-ha wiki
as well as past postings on the mailing list. As far as I can tell,
everything is working as it should with heartbeat. I am trying this
mailing list as I think it has the best possibility of providing me an
answer to my problem.

My config -

HA NFS Server:
2  Servers connected to a single backend fibre array. (Suse SLES 10 /
EMC PowerPath / EMC CX)
Heartbeat controls:
1 XFS filesystem  (/export/test)
1 ext2 filesystem (/var/lib/nfs)
nfsserver
VIP of the NFS cluster.

Client:
Debian Etch w/ nfs client

Now for the problem. I bring up the cluster and everything starts on
node 'a'. I am able to mount /export/test on the client and all works
well. If I induce a failover to node 'b'; all of the services fail
over to node b  and start in the proper order (filesystems, nfsserver,
vip). However, on the client I get the dreaded 'Stale NFS Handle'
error. At this point, I can umount and remount the share on node 'b',
or I can fail heartbeat back to node 'a'. both actions result in me
being able to use the nfs resource.

SLES 10 does not come with a statd daemon, for some odd reason. I am
left wondering if this is the missing piece to my puzzle. At this
point I am beginning to wonder if I should be using a different
distribution, or if there is something totally obvious that I am
overlooking.

Thanks for your time,
Ryan
_______________________________________________
Linux-HA mailing list
[email protected]
http://lists.linux-ha.org/mailman/listinfo/linux-ha
See also: http://linux-ha.org/ReportingProblems

Reply via email to