Volumes out of space

[Rumple]

Here is an interesting and confusing issue I've hit in 2 different environments in the last week which make no sense.

Environment 1 - HyperV 2012 R2 with ISCSI CSV volume. This server is the destination for a veeam replica. One evening the CSV ran out of space (2tb volume) and when I checked the folders the replica VM folder was 1.5TB in size. I added 1TB to the volume and resized disk in Disk manger. Suddenly I have 2TB free in disk manager...I go look at the VM replica folder and its ~300gb (which is the size of the source VM). No veeam jobs were running, it just magically reduced in size.
I scratched my head and ignored it as I could not find any cause.

Environment 2 - HyerV 2012 R2 with NetApp SMB storage for VMs. These VM's are backed up off disk with Veeam. A couple VM's paused with out of disk space errors about 30 minutes after the last veeam backup ran which didn't include those VM's (so nothing was actively running at the time). There were other VM's on the same disk but they did not pause, and there was about 500Gb free on the smb share according to windows. However on the NetApp there was about 3TB of physical space (which would be about 6TB of logical space with inline dedupe,etc)
I decided to try the same trick and resize the SMB volume from 6TB to 6.1 TB....suddenly windows shows 1.5TB free (from a 100mb change).

I've never run into this before and it seems rather strange for 2 in a week in completely different environment setups. I've installed treesize on a VM with a mapped drive to the smb volume and if it runs low on space I am going to run a treesize before and after the SMB resize operation to see if I can figure out what VM/folder "thinks" it is vastly different in size to see if its a particular VM.

Anyone else ever run into something like this. It may or may not be related to veeam, but at moment, outside of windows (and they are different patch levels too), that's the only commonality between the environments I can think of...

[Mike Resseler]

Hi Mark,

I have absolutely no idea what this could be. I honestly can also not tell (or confirm) that this is Veeam related. I have never encountered it on my side.
If this happens again, and you do the before/ after with treesize. Can you also export the logs before and after and maybe create a support case if you see the treesize having data that might be related to Veeam?

Thanks
Mike

[Rumple]

The other environment using hyperv with SMB had it happen again last weekend. The file Share on the hyperv side showed that there was 1.5TB of free space on the mapped drive (this time) but hyperv was complaining of lack of space.
They didn't do a treesize check, but I might have them do the support case and upload logs.
such an odd issue...
I personally was thinking it was something with netapp and remoteVSS making the volume think its full since it wants 100% free space, but since I hit it on the Tegile ISCSI volume where I don't have that issue, it can't be related.
one of those beat head against wall issues I think

[Mike Resseler]

Yes, an odd case.

I do indeed think it is best to do a support case and upload the logs.
Keep us informed

Mike