NW65SP8, NCS1.8.2
Given a two node cluster from 'Node1' and 'Node2'.

Node 1 holds one unclustered volume 'data1' (because this volume will be mapped by some special clients via IPX). There's a CIFS share on that volume which works fine for some other clients (SLES).

Both nodes share two cluster resources 'pool1' and 'pool2'.
Now I have to add another CIFS share to the clustered 'pool1' which is at the moment physically on Node2.

I followed the TIDs and docs (Novell Documentation) and added the extension 'nfapCIFSConfigInfo' to the virtual server object.

Then I added the 'CIFS ADD .cn=pool1,ou=abcd,o=efgh,t=ijkl' and the corresponding unload command to the cluster resource load and unload script.

After that I tried to migrate the resources for the new CIFS commands to take effect. This leads to an comatoese state, I think because of unloading CIFS which is at this moment not loaded for the cluster resource failed.

So I removed the CIFS commands from the scripts and took the resource online again.

No I try at first to configure CIFS via iManager on the virtual server.
If I select the virtual server object, I was told CIFS is not running. Trying to start it has no effect (on the physically machine cifs is running).

If I now change from Tab 'general' to the Tab 'shares' and vice versa, iManager states CIFS is now running on the virtual server object.
Fine. I configure the virtual cifs server with the name 'pool1cifs', ip is
Clicking Ok leads to an error: The parameters for CIFS_SetServerConfiguration are invaild for this server type....
Again CIFS is stated as 'not running'...


1. Is it possible to have two instances of CIFS Servers on one physically machine (One for the unclustered, the other for the clustered resources)?

2. If so, how to start CIFS for the virtual server object? Cifsstart will not work, I guess

3. Am I doing something generally wrong?