VMware Solutions Discussions

Reclaiming space on an NFS datastore (best practice)

JSITSUPPORT
6,278 Views

Hi All,

Could someone please advise what is the current best practice as to reclaiming space on a VMware NFS datastore.

Everywhere I read there seems to be a different answer so I would like to know the best practice, the methods I have seen are,

1) Use storage vmotion (we don't have this but can shut the server down for cold migration

2) Use snap drive (hole punching)

3) Use the VSC to use the "reclaim space" option

4) Using third party tools such as sdrive to first zero the blocks, then vmotion

It seems the easiest way to perform it would be to use storage vmotion (or cold migration in our essentials plus environment) however would this perform the reclaim correctly or would I have to zero the white space aswell?

Sorry for my limited understanding, unfortunately I am not a dedicated storage administrator.

Any feedback much appreciated.

2 REPLIES 2

bsnyder27
6,034 Views

I've done option 4 with success.  

  • zero out the free space within the OS
  • svmotion vm to different datastore
  • svmotion back to original datastore

according to the 'du' command in linux, the blocks are actually freed up again.

 

I still question whether or not this is a recommended practice by NetApp or not though.

shatfield
6,024 Views

on essentials, just zero the blocks within the guest and let the next dedupe run take care of it. 

Public