There are multiple ways to handle this...
My advice would be to run a 3x5 perfstat during the times of high utilization and open a support case with netapp to have them help read the perfstat.
The other ways you can do this is
priv set diag mode and run statit -b - wait 30 seconds, run statit -e
This will give you a lot of information. Especially some of the aggregates and what the disks are doing...
Plus, you can play with the stats command, but you need to know what you are looking for..