[Linux-cluster] Asking information about NFS /CS4 Cookbook

[Alain Moulle]

Hi

In the RH CS NFS Cookbook by Bob Peterson,
I see this in Step 7 :

"In either scenario, NFS clients continue to read
and write files on the NFS-mounted path without detecting when a failure
occurs"

I could understand this when using "Managed Virtual IPs" but
I can't when using "Managed NFS service" :
in this last case, suppose we have a shared ext3 FS1 which
is exported on node1, and so reachable from a client node
which has mounted this nfs FS1. In case of failover, the
FS1 will be exported on node2, so the same client will have
to change his mount to reach the FS1 from node2, how could it
be "without detecting  when a failure occurs" ?
Sure I miss something somewhere ?

Thanks a lot
Alain Moullé