After a failed upgrade to CU6, we downgraded via VM Snapshot restore but unwanted/not previously monitoring processes metrics were enabled for some reason.
The customer followed these steps to rectify the unwanted process probe metrics:
Navigate to OC Inventory->Groups->click on a group->Monitoring Config->Setup_Processes, and then uncheck the unwanted QoS metrics.
Click Save.
Repeat the steps above for each of the other affected host groups.
Wait for all of those profiles to redeploy.
Check again for profile deployment problems.
An updated version of mon_config_service 23.4.6.3 was also developed and used to cleanup the underlying leftover device profiles.
The callback find_overridden_device_
This callback was part of CU8 and recent enhancements to the callback will be delivered as part of CU9.