Hi
@AlexR
Let me know if this helps any:
Basically the difference is that the Controller IOPS is pulled directly from stargate, so that reports what the front end sees. The Disks IOPS on the other hand is pulled directly from the disks, so it reports what the backend sees.
Things like memory cache hits means that the front end may be higher then the backend, and things like read-ahead could make the backend see more reads then the front end.
Also check out the
Nutanix Bible - Search for Controller Layer - you should find additional information. 👍
