I have a question on enabling MonSesCPUNormalization parameter. Currently we have appliance and enterprise platforms.
I know With MonSesCPUNormalization enabled, exception management will trigger based on normalized CPU values that represent CPU usage metrics in terms of a common CPU platform speed. And as per best practices, DBAs should strongly consider enabling the DBSControl Parameter MonSesCPUNormalization, and then determining any CPU time thresholds based on data mining normalized CPU usage metrics instead of regular CPU usage. Without MonSesCPUNormalization enabled, the particular threshold values for the three CPU oriented exception thresholds (CPU Time, Tactical CPU Usage, and CPU milliseconds per I/O) will need to be tuned every time a system upgrade is experienced.
But when we talk about CPUNormalization, generally it is predominant in coexistence/co-residence of different node types systems, but i am not sure whether anyone of you are following this recommendation and enabled this parameter on any environment?
Please share your thoughts or suggestions on this.