VMFS Datastore increases in the middle of the night

[DukeR]

Hi all

We do have a situation which we do not understand yet. In the middle of the night (3 o'clock) one of our VMFS stores gets consumed up to 97% which means a data grow of 1.5 TB within 1 hour! After that the data disappear again.
There is no backup job running or anything else I can see! There is a certain regularity Wednesday, Friday, Saturday, Monday but withdifferent time slots 4, 3, 7:30, 4:40 and 8:30 am

Do you have any good idea on how to determine the root cause?

Based on the monitoring we know that there is no data growth on the VMs iteself so it has to come from outside the VMs. But as I mentioned Veeam seems not to do anything during this time. I still assume tough that it could be snapshots but WHY so big? So you see it is all a littble bit conflicting.

Thanks and cheers

[Andreas Neufert]

Just guessing... Antivirus, Storage reorgnaisation, defragmentation?

Anyway, we as Veeam just create snapshots at backup and delete it afterwards. As it is not inline with your job schedules we do not do anything at VMware.
Potentially our Snapshot hunter delete old VMware Snapshots that hang on VMware side, but this should not happen that often.

I would recommend to install Veeam ONE (30 day trial) and analyse what happens at these times. You can see if the IO comes from within the VMs or are outside of the VMs.
As well you can schedule a garbage file report for these times to see if something else create files on the vmfs.