No... By virtue of the Rename Node, which is the crux of the technique, the HSM (and non-HSM) filespaces historically owned by node_a would be owned by node_b, and so the stubfiles restoral on node_b would be natural, no -virtualnode. It is any later restorals which node_a needed, for data that it had backed up before the rename, that would require it to then pose as node_b, which now owns the historical filespaces. Any new node_a backups would generate fresh filespaces under its node_a name.
You may have some test systems where you could give this a try, to get familiar with it, and test its full effect, first.
Richard Sims
On Feb 23, 2005, at 5:45 PM, Michael Prix wrote:
If node_a and/or node_b already have other filespaces, you could consider employing -VIRTUALNodename for restorals of previously backed up data.
So, as this will be the situation: a) create a filesystem b) start HSM on that filesystem c) do a restore of the stubfiles for this filesystem with -virtualnode=node_a
