What would you do to get a full picture of CPU utilization on the entire Exadata RAC cluster?

The mpstat output from OS Watcher shows a database node as being 90% idle on an average.
What would you do to get a full picture of CPU utilization on the entire Exadata RAC cluster?

A.
Average the mpstat id1 output from all the nodes.

B.
Ask application users if they have noticed a slowdown in screen response.

C.
Look for an increase in batch job servicing times.

D.
A & B above

Explanation: