High CPU IO Wait

[WillemDH]

Wednessday we migrated our Nagios XI production server to another VM datastore. This datastore is almost identical to the original datastore.

Since the migration the IO Wait of the server statistics sometimes goes into 'red'. Before the storage vmotion, I never saw it go into yellow.
See attached screenshot for a view of a top during the high iowait.

So my questions are: Why do we get such a performance difference when switching to another datastore? Can this be related to the backend storage system? Or is this purely a cpu issue? The system is still running on the same esx server and has 6 cpu's. So load is not the problem, as you can see in the attached cpu load graph. The server has 4 GB RAM, but ram does not seem to be the issue.

[WillemDH]

Hello,

I found out that there was a difference between the old and the new datastore. it seems the new datastore is a T2 with SAS & SATA, managed by Automatic Storage Tiering (AST). it seems ,as the vm has been moved recently it could be on low tier storage. I've convinced my boss I need T1, which is SSD & SAS. (also with AST)
So Please keep the thread open for a week or so and i'll let you know if my io wait problems are solved.

Grtz

Willem

[tmcdonald]

Will do.

[WillemDH]

IOwait is gone since we moved it back to SSD / SAS luns. Thread can be closed. Tx