Nimble showing less usage than VMware on Datastore

[TrentW]

We have a HF20-2P-21T-2880F with a total of 15 TB. On our array, our total usage is: 10.3 TB of 15 TB, resulting in 4.7 TB free. When I look in vCenter on the same datastore, it shows: 1.98 TB of 15 TB. I've seen some other threads similar to this, but I'm just confused as to why VMware can't update its available space according to Nimble. If this is a VMware constraint, then why even compress the data when you cannot use the free space on Nimble?

If anyone can shed any light on this, that would be perfect and I appreciate any help.

Thanks!

[msobers22]

Hello,

Let me first make sure I understand what you asking. In the Nimble GUI you see the available free space 4.7TB and in vSphere you see 1.98TB free space for the datastore.  If this is the case then I believe the issue might have to do with UNMAP or Space reclaimation on VMFS Datastores. See the link below to see if it matches your issue.

Space Reclamation on VMFS Datastores (vmware.com)

Hope this helps,

[TrentW]

@msobers22 

I already have space reclamation setup on this datastore. It is configured as: "Enabled at fixed rate", rate: "100MB/s". Similar to the guide you had sent. Do you suppose it's not working automatically? I will attempt to manually run the command and see if that does anything.

Thanks!

[Thomas_Lam_HK]

Do you having any storage snapshot on that datastore? 

Having storage snapshot will locks the blocks underneath which is the most common reason blocking the UNMAP process to reclaim those space, as long as the snapshot is retented and no more pointer is associating, free space can be reclaimed.

[somoa]

I had same happen to mine, support told me to do a controller failover. It cleaned up the unused space after this.

I was told this restarts/triggers the garbage collection process.