reverse incremental strange behavior

[cgeorge]

I have a backup job off 3 vm's that starts at 5h
I did a full backup on a saterday off those vm's and afterwords reverse incre with retention of 7.
The strange thing is that every saterday the backup jobs takes 6h and all the other days 1h30. It's like he is taking a full backup, but this is not configured?
Any idea

[emachabert]

If no weekly full is planned, you should look at the counters : read and transferred.
Perhaps, there is huge workload Inside the VMs on saturday that changes a lot of blocks, and then CBT journal is huge.

[cgeorge]

I've checked the read and transferred, there almost the same for everey day?

[emachabert]

Ok so if the counters are the same, and the duration is not, then there is a slowdown on saturday. What's the throughput on saturday compared to other days ?

[foggy]

Are there probably any other tasks targeting the same repository at the same time? What are the bottleneck stats for both job runs (fast and slow)?