Swap file blocks skiped, only a fraction of swap file size

[Aurock]

When I look at the backup statistics of my main backup job and select a single vm in the list, I see a line in the right pane which says "Swap file blocks skipped", followed by a number of MB or GB. This number seems to be significantly smaller than what the actual swap file size is on the VM. Why is that?

I assumed maybe it was because some of the blocks (in sizes veeam looks at them) might contain data other than just swap, but in one particular case, I know that the swap file is on a separate VHD, with nothing else on it. The "Swap File blocks skipped" for that VM says 1.4GB, when the actual swap file size is 32GB. Why wouldn't veeam be skipping the entire 32gb of data, effectively the entire 2nd vhd file?

In other cases, the difference isn't quite so dramatic. Most of our servers don't have 32GB swap files, but on some of those servers Veeam says it's skipping as little as 2.8MB of swap file blocks.

[foggy]

I assumed maybe it was because some of the blocks (in sizes veeam looks at them) might contain data other than just swap

Correct, as Veeam B&R operates by 1MB blocks, some of them could contain also other data due to disk fragmentation and should still be fully transferred to target (hence, those are not excluded and are not accounted in the displayed amount of skipped swap file blocks).

but in one particular case, I know that the swap file is on a separate VHD, with nothing else on it. The "Swap File blocks skipped" for that VM says 1.4GB, when the actual swap file size is 32GB. Why wouldn't veeam be skipping the entire 32gb of data, effectively the entire 2nd vhd file?

Other factor that should be considered is whether the particular job run is full or incremental. During incremental job run, swap file blocks to be skipped are excluded from the blocks that have changed since the last backup and obviously not the whole swap file is typically changed between job runs.

Besides, blocks that belong to swap file but are already marked to be skipped during previous data processing stages (zero blocks, for example) are also not accounted in the displayed amount of skipped swap file blocks (though they are accounted in the actual swap file size inside the guest OS).

Hope this makes sense.

[Aurock]

Correct, as Veeam B&R operates by 1MB blocks, some of them could contain also other data due to disk fragmentation and should still be fully transferred to target (hence, those are not excluded and are not accounted in the displayed amount of skipped swap file blocks).

Okay, I expected as much. Though in our case, I have the backup job set for WAN target, so that would still apply, but in 256KB blocks instead of 1MB, correct?

Other factor that should be considered is whether the particular job run is full or incremental. During incremental job run, swap file blocks to be skipped are excluded from the blocks that have changed since the last backup and obviously not the whole swap file is typically changed between job runs.

Besides, blocks that belong to swap file but are already marked to be skipped during previous data processing stages (zero blocks, for example) are also not accounted in the displayed amount of skipped swap file blocks (though they are accounted in the actual swap file size inside the guest OS).

Hope this makes sense.

Yup, makes perfect sense. Thanks for taking the time to explain it.

[foggy]

Okay, I expected as much. Though in our case, I have the backup job set for WAN target, so that would still apply, but in 256KB blocks instead of 1MB, correct?

Correct.