is it true that pseudo device major number of a powerpath device must be the same on all cluster nodes to properly handle NFS mounts?
is it not enough keep shared volume group's device major number the same on all cluster nodes? (the one exported filesystems comes from)? this is what I have practiced till now...