We have a FAS3140 with a 3 TB volume with 3 1TB LUN's in the volume. The LUN and volume are thin provisioned with space savings being pushed back into the aggr. I sent my ESX admin a DF on his volume because he said he is running low on space. Why is ESX reporting that he only has a 200 GB in the datastore to create additional VMDK but the array is reporting 1831 GB? ESX server is 4.0 and they use thick VMDKs?
Filesystem total used avail capacity Mounted on /vol/esxvol1/ 3072GB 1240GB 1831GB 40% /vol/esxvol1/
LUNs that are deduplicated cannot realize the space savings gained by deduplication. LUNs are a logical presentation of a SCSI disk drive, which have no correlation between the LUN's formatted filesystem and WAFL. If you want your ESX admin to see the deduplicated space, migrate your VMDKs to NFS if possible. NFS datastores are WAFL so the space savings from deduplication can be realized by ESX.