ESXi datastore configuration

[usao]

I have an application where I need to sustain 2+ GByte/sec throughput and im finding that the default datastore configuration is not supplying anywhere near that throughput.

Im looking for a way to create my datastore as a striped configuration on 4 luns from 2 different EMC arrays accross a set of 4x 16GB FC ports.

This will give me the required bandwidth, but im not finding how to setup the datastore as a stripe of luns.

Also, it would be very helpfull if I could resize the datastore, by adding/removing luns, but again im not finding any way to do this.

Can anyone point me to the instructions on how to stripe, add and remove luns from a datastore?

Thanks

[parmarr]

You can see https://www.vmware.com/content/dam/digitalmarketing/vmware/en/pdf/whitepaper/vmware-vsphere-vmfs-bes... - even though it's for vSphere 5.1, it should still apply. Also, see https://www.vmware.com/content/dam/digitalmarketing/vmware/en/pdf/techpaper/vmware-perfbest-practice.... Has the customer also check if there is SAN best practice documentation specific to his hardware?

[ITaaP]

Take a look at this. VMFS Extents - Are they bad, or simply misunderstood? - VMware vSphere Blog

Also, from VMware documentation.

Additional capacity with disk spanning does not necessarily increase I/O throughput capacity for that VMFS

volume. It does, however, result in increased storage capacity. If properly provisioned on the underlying storage

array, the additional capacity can be allocated on LUNs other than the first LUN and will result in additional

throughput capacity as well. It is very important to be certain you are adding LUNs of similar performance

capability (RAID level and I/O density) when adding to an existing VMFS volume.

You may want to speak with your EMC VAR to see what is the best way to achieve the performance you are looking for.

[dekoshal]

If you found this or any other answer helpful, please consider the use of the Correct or Helpful to award points.

Best Regards,

Deepak Koshal

CNE|CLA|CWMA|VCP4|VCP5|CCAH

[usao]

Ive achieved the necessary performance when I stripe luns on the host side, I just can't find a way to get the datastore striped.

When I map the devices as RAW devices to the VM, I can achieve the necessary throughput, but when I use datastore devices I cannot achieve the necessary throughput.

So, my solution has been to use RAW devices and stripe at the VM layer.

Thanks