V11 Upgrade - some VMs disks are read completely

[mkretzer]

Is it normal that some (random) VM disks get read completely after an upgrade (V10->V11 in our case)? CBT is shown besides the disk, nothing about "active full" or "CBT reset" - it just reads the disk completely one *some* but not all VM disks. And only for the first backup run! Interestingly, it happened on a machine with nearly no changes (16 TB processed, 7.5 TB read (5 of 13 disks completely), 7.6 GB (Giga!) transferred.

I have seen this with previous upgrades but never opened a case because it always went away after the first successfull backup...

[Gostev]

Yeah I vaguely remember devs telling me something about this, basically this has to deal with fixing/upgrading auxiliary file system metadata in backup for the BitLooker functionality that allows us to drop blocks of swap and other useless files, blocks belonging to deleted files etc. Sometimes following the new version install this metadata needs to be refreshed or updated depending on the detected disk configuration. But this should only ever happen on Windows VMs, and never for other OS.

[mkretzer]

Sounds about right. I have only seen this on Windows VMs. Thank you!

[mkretzer]

Hmm. Most of our MS SQL Server are affected. Backups are still going...
This might be something for the release notes so you can prepare...