Storage performance monitoring help pls.......

[zenomorph]

Hi there,

Were hitting an performance issue on our 4.1 ESXi Cluster issue with storage. We are using EMC Clariion CX3-80 RAID5 and have some metaLUN's with datastores and some VM's running with RDM. What were finding on these RDM's is very high "disk queue length figures" showing on both the SAN LUNs and the VM's themselves.

Also what were seeing is on the ESXi's were seeing the "Storage Adapter" read latency max. at 15ms and averaging at 4ms and on the VM's with the high queue length were seeing read latency of max. 33ms and averaging 3.1ms. Can I ask are these figures considered high, on the OS itself were seeing disk queue length of 4-5.

Any suggestions on what the cause might be, were running Domino mailboxes on these VM's and were using RAID5, would changing it to RAID10 fix or any other suggestions?

Many thanks.............

[idle-jam]

have a look at this and it will tell you the threshold .. http://www.yellow-bricks.com/esxtop/

[FredPeterson]

The type of RAID isn't nearly as important  as the number of spindles.  A RAID 5 can be 3 or 5 or 10 or 20 spindles.

Disk queue lengths, at least in the physical world, should never exceed the number of spindles.  Latencies should rarely exceed 30-50ms for any substantial length of time.  There is a reason VMware chose 30ms as the SIOC threshold - which you should enable if you can.

Are you actually experiencing what you believe is a performance bottleneck as a result of disk?