Hello Jon,
We did do all those things you suggested but support was not able to pinpoint the problem.
Thanks for the tip Frank. We might have to use that if it pops up again. At the moment things appear to have stablized. It would appear after upgrade the machines were kicking in huge amounts of error scans and out of sync scans which were forcing full scans. We had to basically brute force our way through the performance issue by maxing out our DB threads for the inventory service and let it run badly for about a week.
Just so everybody knows our hardware that runs our LANDesk environment should be overkill. It is a core server running 32GB of RAM talking to a dedicated DB instance that has 60 GB of RAM on the same switch. We are managing 15K nodes on the core server that was most problematic.