Clusters can do weird things sometimes... have you tried explicitly specifying the volumes for physical and virtual nodes and seen what happens?

Yes, we tried saveset All and explicit volumes. We tried backup command
save and save -c virtnode. No difference.

This is known issue - contact support for patch.

I'm not going to pretend to be an expert on Netware (something which salespeople have tried doing on my behalf in the past ) but this may be the expected way of backing up the Netware cluster volumes. With other operating system clusters you backup data from the virtual node use save -c virtualhostname, then when you do a recovery you do it from the physical node directing it to the virtual node.

[STRIKE]This might be an problem the way you load your TSAFS.NLM
There are parameters like NOCLUSTER

If I remeber corectly you have to load the TSAFS.NLM like:

TSAFS.NLM /NOCLUSTER[/STRIKE]

I have to read better next time.
I second Hrvoje, its an known issue.

Message was edited by:
Antoine Schokker

Message was edited by:
Antoine Schokker

Message was edited by:
Antoine Schokker