incremental backup very big with few fliesystem change

[ma7u]

Good afternoon,
I have a problem with an incremental backup of a Linux Debian 9.0 machine hosted on a Hyper-V.
On the target machine, there isn't any veeam agent installed.
I do VM full copy incremental backup daily with a synth full backup on Saturday.
I was monitoring the size of the incremental backup and I can see from the report of veeam that every day the incremental backup is around 20GB, the full machine HDD is 100GB where only 50GB are used. I checked with df commend the ext4 filesystem every day and there are just a few MB of difference.
i attach here the report of Veeam.
Could you please someone help me to understand why this incremental backup is so big compared to the small modification of the filesystem?

thanks

m.

[Dima P.]

Hello ma7u,

Incremental backup size reflects not only new data but changed data since last backup job run as well. Aside from operating system files which could be modified by new packages, is there any application installed on the machine that might modify files on the file system which affects the incremental backup size? Thanks!

[ma7u]

The Linux server is a webserver hosting one website. PLESK is configured.
No change are made, daily, to the file system.
Is it possible to check what file exaclty veeam if considering changed since the last 24h?

thanks

[PetrM]

Hello,

I don't think that the issue happens in the guest OS but most likely at the level of hypervisor. For VMs running on Microsoft Hyper-V Server 2016 or later, Veeam leverages native Microsoft change tracking mechanism - Resilient Changed Tracking. In short, this mechanism finds changes between the reference point (that is created after a previous checkpoint is merged with the base disk) and the current checkpoint. I assume that in your case, there are lot of changes from RCT perspective. If you have enough storage for an extra full backup, perhaps it would make sense to check whether the issue reoccurs after full backup. Also, I believe you should start investigation with our support team so that they can validate my assumption. Please don't forget to share case ID with us.

If you run the workloads on Hyper-V 2012 R2 or earlier, you can try to reset CBT according to this instruction.

Thanks!

[ma7u]

mea culpa.
was a script of backup on the server that was creating every day a 20GB backup new file, that veeam was incrementally backup (correct)

SOLVED.