Due to bug CSCtu18692 I need to increase the vCPU commitment on a CUCM 8.5 VM. The publisher is pegging CPU every night during DRS and TAC's guidance says it could cause outage and should be addressed.
I've read mixed results on adding vCPU resources to a Linux VM. Some suggest a rebuild.
VMware specifically says when doing this to Linux boxes you should be rebuilding the OS. Now with that being said I have done it in the lab a couple of times now, since I have ran into the same issues. A couple scenarios I noticed when doing this:
-One VM's services had issues starting after the reboot
-A couple ran for a couple weeks with moderate use and then unexplainable things happen to call processing
-Have had a couple VM's that havent had any problems YET....
Personally my recommendation would be to take a backup of the database and rebuild the VM. By doing this you could save yourself hours of troubleshooting down the road for something that may not even be able to be explained.