Ran out of disk space - options?

[evander]

I have a replication job that is quite big. My Datastore is 1.4TB in size and I dont have any more available disk left on the Datastore. The entire Datastore is dedicated to this replica
The replication initially worked fine for the first two days or so but then ran out of disk space. Currently my Veeam Proxy box (also virtual machine but on different Datastore) is hung with the normal Vmware error OutOfSpace and click Retry or Continue.

Question 1:
Can I manaually delete all or some snapshots on the replica to free up some space and unfreeze the Veeam Proxy server?

Question 2:
What is the best way to minimize the amount of space the replica will take up in future?
I'm thinking possible answers to this would be to bring the number of restore points down to 1 (1 is better than none)
and obviously try and delete any unwanted data on the actual server.
Will changing the deduplication settings to maximum help all that much? I am always a bit weary of the processing overhead. Its currently set to Lan Target

Thanks,

[lloyd.mills]

Hello evander,

If you power off your VM you should be able to remove the snapshot.

What datastore is actually full?

It sounds like the datastore which the Veeam Proxy is on is full, which should not get affected by the snapshot of the VM's you are replicating. Have you identified which server/s have an open snapshot/s?

[foggy]

Yes, you can safely delete some of the replica snapshots using vSphere Client (though be noted, that corresponding restore points will be preserved in the database until cleared up according to retention settings). And you are right with answers to your second question (number of restore points and clearing up the VM will do). As for deduplication/compression, they only affect the VM data in transfer between sites (to reduce network traffic). The replica VM itself is stored in native (undeduped/uncompressed) format.

[evander]

Hey Lloyd, thanks for the reply. So here's the strange thing. The Veeam proxy and the Replica both sit on the same Storage but different Datastores (both on ESXi)

The Datastore holding the Replica is only housing the replica and that is now full with no diskspace available
The Veeam proxy server (windows 2008 R2) is sitting on a Datastore with plenty free space but it's half way thourgh a repliaction job when it failed and it has frozen trying to create the snapshots (I assume)
If I look at the harddrives assigned to the Veeam Proxy box (besides the C and D drives) there are other drives which point to the Replica Snapshots on the Replica Datastore. This would release itself I am sure once the job finished...but it never did.

So I am hoping I can simply through Vcenter delete the first or more snapshots assigned to the Replica and in doing so free up space and let the job continue???

[evander]

Hi Foggy, just saw your reply after I replied to Lloyd.

So I deleted two snapshots and unfortunately it didnt recover any space on the datastore which is quite strange. Vmware also completed the delete process extremely quickly which was unexpected. Even after refreshing the Storage views its still hasnt reclaimed any space.

The only Snapshot I have left is the "Veeam Replica Working Snapshot" - can I also delete this though Vsphere Client??

[foggy]

This one is a helper snapshot used for replication itself. You can delete it if the job already managed to fail and the snapshot still resides on the storage. You can check what files are actually taking space by browsing the datastore using Veeam backup server UI.

[evander]

Thanks Foggy - I ended up deleting the snapshots and the entire replica too as it wasnt able to free up enough space. Word of advise for anyone having the same problem, remember that even after I deleted the replica Virtual Machine through Vphere Client I still needed to manually remove the 3 harddrives listed in the Veeam Proxy which were pointing to the snapshot files in the Replica Datastore. They had not released yet because the job failed halfway through with the diskspace issues. After I edited the setting of my Veeam Proxy server and manually removed those hardrives pointing to the Replica everything was gone and the datastore had reclaimed all its space.
So basically I will start from scratch which is fine for now. I have scheduled the big initial replication to happen over the weekend because its still over 1TB in size.

Thanks for the help.