Corrupt/Missing disk

[meeyou]

Attempting an FLR and finding that the top level folder that contains the desired file is showing empty in the FLR browser.

When I check from windows explorer I first get a permission error, when I tell it to take permission I then get an explorer error that the file or directory is corrupted and unusable.

I tried a chkdsk /f on the mounted FLR volume from the Veeam console machine, and this allows me to explore the tree, but the subfolder containing the desired files doesn't exist. When I check this folder on production, it was created 2 days prior to the restore point I'm looking in, so it should definitely be there.

I've checked several other restore points in that chain and they all show the same behavior. Job history is error-free for all of the points I've checked.

The desired file should also be available via shadow copies in the most recent chain, but in this chain the disk just doesn't show up at all, either in FLR browser or windows explorer (we have one fewer disks than expected when browsing the FLR mountpoints).

I just opened case 03106075 a few minutes ago, waiting on my SWAT callback currently, so I figured I'd post it up here for more input. If we can't restore this file that's one thing, but it's concerning that the issue seems to have gotten worse in the most current chain.

VBR 9.5.0.1922, Hyper-V 2016 host/VM, Server 2016 guest, Server 2012R2 Veeam console(this disk is not ReFS)

[Andreas Neufert]

Try to install Veeam Console on Win2016 or 10. Maybe your 2016 server use a feature that is not compatible with Win2012.

[meeyou]

Well working with support we confirmed that the disk is just not backing up completely/usably during active fulls or subsequent incrementals, even though it's reporting success in the job.

We're currently attempting replacing the disk with a new one and migrating data from one to the other in-guest. I'll be looking at the next backup run to see if it copies a reasonable amount of data for this disk.

It's a 1TB dynamic disk that's expanded to 550GB, with 505GB used, and in active fulls it's copying ~16-25GB, and the disk is inaccessible in any of those chains.

I'd like to have a defect to track and perhaps a method of scouring our backup logs for large disparities between disk used bytes and disk copied bytes in job logs so we can identify if there's any other disks affected in this way before we're attempting to recover data from them.