topic space reclaimation not working in StoreVirtual Storage

space reclaimation not working

yann2 — Thu, 16 Jun 2016 20:36:50 GMT

Hello,

I have 4 hp lefthand P4330 on 12.0..00.0725.0 with volume using network raid 5

As I understand it, from version 12 and up, I am able to use space reclaimation.

however it does not seems to be working, I tested it with a datastore that have those detailed:

vmware report:

capacity: 2tb

free space: 1.7tb

use: 300g

hp p4330:

reported size: 2tb

consumed space: 922.88gb

provisionning: thin

protection level: network raid 5

by using the commandon vsphere 5.5:

esxcli storage vmfs unmap -u *volume_id*

I was expecting to free about 600g, however nothing change on the p4330, I send the command 1 time and let i run then resend the command the day after, to make sure it run, by using esxtop, it seems to be running on the vmware.

but the reported storage on vmware and the lefthand have not change.

any help would be appreciated, we are losing alot of disk space because of this problem, in fact this is my smaller datastore...

Re: space reclaimation not working

oikjn — Thu, 16 Jun 2016 21:21:52 GMT

DId you enable sapce reclaimation? YOu have to right click on the management group and enable it there first.

Then I can't remember how you enable ESX to see it for existing LUNS, but the easiest way to test is create a new LUN and connect it after it is enabled and it should function.

Re: space reclaimation not working

yann2 — Thu, 16 Jun 2016 21:34:52 GMT

yes space reclaimation is enabled on the san group in the CMC

the lun were already connected to our vmware host before the space reclamation was enabled:

heres how it was done:

lun connected to host -> update p4330 to version 12 -> enable space reclaimation -> run vmware command :

esxcli storage vmfs unmap -u volume id

Re: space reclaimation not working

Stor_Mort — Fri, 17 Jun 2016 15:02:37 GMT

After enabling space reclamation on the StoreVirtual management group, it is necessary to rescan all the volumes on your VMware hosts so they recognize the status change. Then try the unmap command.