Customer has a Celerra NS-120 supporting VMware data stores on NFS.  The underlying file system is created on an AVM-provisioned R1+0 volume spread across a number of 2-disk R1 volumes on the underlying CX4. 

This datastore experiences severe latency issues, and I believe it is because the Celerra has put most of the operational data on just a few of the underlying disks.  i.e. the I/O workload is not being evenly distributed across the underlying d-vols.

I'd like to work with the customer to tear down the file system, and re-create the d-vols on native R1+0 provided by the CLARiiON, to get a true workload distribution. across the disks.

Looking at the documentation, is seems that native, more-than-two-disk, R1+0 is not supported by the Celerra.  However, I thought that this had been supported in later DART releases.

Can someone clarify?

Thanks,

Eric