Thanks heaps,
It does seem a bit too complex for what it achieves as you point out,
Could I ask your opinion on the attached?
- basically add another NFS store, have 2 datastores on each with each host using a different store and replicating each store to the other.
Still a manual process to recover from a Datastore failure but I can handle that.
I think this would provide better performance under normal operation too and hopefully simplify the setup a bit.