NFS datastore ran out of space, can you tell which VM was the cause?

[Guv]

We are using vpshere 4.0, and we use some NFS datastores and some VM's have thin disks.  It looks like one of the NFS datastores ran out of space, as there are several VM's on this NFS datastore were down, and has the answer virtual machine question.  I answered the question, and now the VM's are up and also increasesd the NFS datastore.   But is it possible to work which VM on this NFS datatstore grew and caused the space issue, its difficult to see.

Any ideas.

[memaad]

Hi,

Check if any VM has snapshot , which will grow to size of base disk and occupy sapce on NFS datastore.

Mohammed

[a_p_]

On NAS/NFS system it's usually the storage itself which dictates the format (thin or thick). With thin files you cannot really tell which VM caused the issue in the end, because all of them could have grown and caused the issue.

André

[EdWilts]

This problem gets especially hard if you're doing any post-process de-duping on your NFS datastores (typical NetApp scenario). A guest may not use any more space but because it's overwritten a bunch of data, the de-dupe ratio drops and you run out of space.

I don't have any good answers yet but I think many of us have seen the problem.

[vGuy]

On a different tone, I would recommend you to use the below 2 vCenter alarms for the thin-provisioned datastores to avoid running into this issue again:

Datastore Usage on Disk AND

Datastore Disk Overallocation Alarms.