Is it possible to delete a CUPS server from a CUCM cluster configuration via CLI? My customer had an 8.5.2 CUCM cluster that included a single CUPS node. CUPS was not used and the VM had been powered off for some time. The CUCM cluster was recently upgraded to 10.5.2 however CUPS was not upgraded. The CUPS server does not appear in CM Administration under server, CM or Presence views. The CUPS reference is still listed in the OS Administration Cluster View. I would prefer to simply delete the object, if possible. The CUPS server was originally built using an incorrect OVA and, should they decide they want the functionality, it would be preferable to just build a new IM&P server based on the correctly sized OVA.
I had quite similar problem just today.
The link below helped me. Try it.
Thanks. I had read that but unfortunately it doesn't apply. The node has to be visible in the CM Administration GUI for those instructions. In my case it isn't. I'm going to upgrade the CUPS server to IM&P v10.5.2 and then remove it properly.
I'm thinking of a workaround guys; can we try adding the CUPS server as IM&P under server list with the same IP add/ hostname, then delete it.
Interesting thought. I'm not sure that would be successful given the 2 major rev difference between CUPS and CUCM. At this point I'm running the upgrade to IM&P. It should register as an IM&P subscriber node in the CUCM cluster at which point it should be removable through the standard, documented process. Bottom line, don't assume the prior contractors cleaned up after themselves.
Thanks. Again though, I was looking for a way to delete the node via CLI. As mentioned in my original post, the node is not visible at all through the CM Administration GUI, which includes the Presence Redundancy Group. It appears to be a stranded object, visible only through the OS GUI cluster view. That GUI does not include a mechanism to remove the node.
I'm suggesting that to avoid the other option which can be done through Root Access into Publisher server as we need to delete it from the ProcessNode table. You may raise a Cisco TAC to help in generating the root credentials.