The Environment of discussion is VMWare ESX 4.0 Datastore on NetApp, of cause

[mwilliamsm]

No snapdrive for Windows was purchaed.

The Luns Reservation set to zero

Volume Fractional Reservation set to zero

Snap Reserve are set to zero or disabled.

Volume is Guaranteed, Autosize, and AutoGrowth are enabled.

The statemt is, if one lun per volume is guarantee (guarantee=volme) and the volume has unallocated space, that “free space,  ” is used for the lun as it is modified and the old blocks are held in the snapshots. But I set the snap reserve to zero.   I did not set the guarantee to “none” (thin provision the volume), because the “free space” will be used by other volumes on the aggr in question and the thin provisioning needs to be monitored carefully as to not run out of space on the VMFS datastore. That would be a bad thing.

The question: If these statements are true, can someone tell and/or show me how to use the reclaimation cli via ssh to reclaim the VMWare (Datastore) lun per voluem free space?

[ekashpureff]

Michael -

I'm not sure what it is that SnapDrive is doing under the hood when you do a space reclaim, but you may want to try running reallocate on the LUNs in question to reclaim freed space within the volume.

(I'm going to try this out in my lab environment...)

BTW Snapdrive isn't that expensive - it's well worth it in SAN implementations.

I hope this response has been helpful to you.

At your service,

Eugene Kashpureff

(P.S. I appreciate points for helpful or correct answers!)