MSA 1000 Average Queue Depth

[2044418Puts]

Hi,

We are experiencing slow MSA performance. I've ran the performance monitor for one minute (see attachment).

Our MSA 1000 consists of two cabinets (14x 300GB and 14x 150GB), redundant controller and redundant 8 port SAN switch.

The MSA1000 is connected to an ESX cluster consisting of two vSphere hosts. Their VMs are running SQL, which are going slowly.

I'm not a storage specialist, but according to the output, the MSA is busy 95,6% of the time just performing read and write actions. I guess this is way to high.

The command latency is 4.66ms.. Is this normal?

There is an AvQDep of more than 300.000 for LUN0. I don't know what this means exactly and what a normal value should be. But if this is the same like the Windows Perfmon queue depth I can imagine that this is a huge queue. What are normal values here?

Thanks for your help.