Performance monitoring of EVA 3000

[Hussain Faiz]

During the checkup, we found that the Total Read Hit Req/S is less than Total Read miss Req/S i.e. 166 vs 747. This means the large chunk of the data is fetched from Drives instead of Cache. The Cache is enabled on both the controllers.
What could be the reason?
This Array hosts the Exchange applications. The Exchange Admin is complaining about the latency in the Application. Could this be a reason for the latency?
Could someone help me in calculating the IOPS for this Array.
This is an EVA 3000 with 24 drives on RAID 1 for Exchange. Each drive is of 300 GB RAW and 280 GB formatted capacity with 15000 RPM.

Thanks in advance

[Víctor Cespón]

Yes, 82% read miss means that the requests are too random and the controller cache seldom contains the data requested.

24 15K drives in RAID 1 can give you 2.748 IOPS, assuming a typical 60% read/40% write. You need to compare that to the total operations being done on that disk group.