We have FAS3240 active/active filer, each with SAS and SATA aggregates. Both controllers are currently in 8.1.1. On our SAS, we have Exchange, SQL, Sharepoint and NFS datastores for VM. We also have a PAMii card on each controller.
Our Exchange server(2007, OS is a VM, iscsi luns for the dbs and logs) latency goes up (as detected by Windows perfmon) most of the time when a vmotion is happening in the VM environment. I monitor latency of LUNS and the containing volumes but normally they won't spike up.
It is really hard to figure out if it is the Exchange, or the VM environment or Netapp storage. I am wondering if we are using our Pamii properly.
Here is how it is configured right now.
If we need another Pamii, i want to figure out what numbers I can look at so I can justify it.
Non disruptive to change. But ask why it was set. Not a default but they may have had Analytics showing it would help. It may cause higher evicts of no read after wrote and would typically see it off. Is it the same on both nodes?