Hello!

We have a cluster that provides a HA-NFS-Server. Nodes in that cluster also 
mount some of the directories exported via NFS (e.g. /foo is mounted from 
server:/exports/foo).

Naturally one of the cluster nodes can be NFS-Server and NFS-Client at the same 
time.

The setup is a bit tricky, especially if you use waiting (non-bg) mounts in 
fstab when the NFS-Server is down (corosync (and the NFS Server possibly) would 
be started only after all the fstab mounts are complete).

With some care, such a setup works (preferrably in a >2-node cluster). In out 
older HP-UX system such a setup also worked, and was explicitly supported as 
"NFS crossmounts".

Unfortunately there was an update to SLES11 SP2 release notes recently that 
forbid this kind of setup:
---
11.5.3.2. Mounting NFS Volumes Locally on the Exporting Server

Mounting NFS volumes locally on the exporting server is not supported on
SUSE Linux Enterprise systems, as it is the case on all Enterprise class
Linux systems.
---

Can anybody explain the rationale behind? (Don't tell me knives are forbidden, 
because you can cut yourself in the finger unless very careful...)

Regards,
Ulrich


_______________________________________________
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