It's possible that what you're experiencing has nothing to do with any of what I'm saying, but you should note that havoc may still be wrought with no RAM or CPU redline manifestation at all (even on a beefy server). If there are db locks happening because the PM db isn't playing well with others (sharing the instance), or if PM is being bullied by other dbs in that instance, this could be an issue.
Is ProcessManager sharing an instance with other DBs? The Symantec_CMDB database is a bad pair for PM for the same SQL instance.
Again, it could be something totally unrelated, but this is how I've fixed many environments with problems similar to what you're seeing.