ReFS FastClone after GuestVM Resize

[parneye]

After resizing one of the backed up VM volumes, from 1.7TB -> 1.74TB, the amount of HDD that was cloned dropped significantly.

Usually with this VM, ~1.7TB is cloned, but after the disk resize, the newest VBK took up a fresh ~1.7TB, only cloning about 25gb.
I verified this using "blockstat - http://dewin.me/refs/"

Is this known behaviour with ReFS, and is there a way to resolve?

It appears that every time we need to resize this large file server volume by any amount (40gb in the above example), we end up using another ~1.7TB in storage on the Veeam ReFS volume, where usually, a VBK for this VM takes up about 20GB. (verified with "blockstat - http://dewin.me/refs/")

This is a Veeam Agent job managed by Veeam B & R, with individual VBKs per VM.

After the resize, the Veeam job detected the change with this warning "Disk 1 layout change detected, resetting changed block tracking for all volumes on this disk"
It successfully completed the fast clone synthetic full: "Synthetic full backup created successfully [fast clone]", but as mentioned, cloning was extremely minimal.

I'm not sure if this is different for HardDrive size changes vs Volume Size changes. In this case, we have unallocated space on the Harddrive that we use to resize the volume by extending it.
In other words, the Disk has 270GB unallocated, we resize the volume by using part of this unallocated space.

Any ideas?

Thanks

[isusko]

Good day James,

I am not a Veeam engineer nor expert, but I guess what happened here is that the geometry of the disk changed when you resized it, causing a minimal cloning. Any ways I suggest you opening a case with the Veeam engineers to check what really happened.

Best,

Isusko

[parneye]

thanks Isusko.

Will open case now.

[parneye]

Case #05192697