We are running SCVMM on a Cisco UCS with a FAS3220 with both shelves filled there a mix of HDD and SSD in each so we have 2 aggregates one on each shelf utilizing flash pools. Each aggregate is approximately 17TB.
We have created a bunch of LUN's from within SCVMM utilizing the SMI-S provider and they are mounted as clustered shared volumes but they only seem to add up the approximately half the capacity of the aggregate yet we cannot create any more?
I am sure there must be a valid reason for this but I don't seem to be able to find it!
I have not used the SMI-S plugin but it would appear that the volume created is sized according to traditional thick provisioning rules. LUN size x2 +snapshot space. So for a 1TB LUN you need 2TB volume plus snapshot space which is typically 20% leaving you with a 2200g volume. This guarantee's space of the entire LUN and 100% overwrite and snapshots.
Have a look and see if you can configure the volume provisioning settings. I suggest reading up on thin provisioning and deciding if you prefer to use that method as it will save space within the aggregate. Just remember if you plan on overcommiting your aggregate you need to have good monitoring in place.