Increasing parallelism for "Applying retention policy for ..." tasks in replication jobs

[DonZoomik]

They are limited to 4 concurrent tasks and it seems to be not controlled by MaxConcurrentDeletingSnapshotsForHost and MaxConcurrentDeletingSnapshotsForCluster registry keys. As it is basically snapshot removal, I'd guess that they are related...
I have jobs with hundreds of VMs and SANs seems to be running quite idle during snapshot removal, dragging out the process.

Is it possible it increase the limit of 4? Some registry key?

[foggy]

Should be controlled by the mentioned values. By default, the number of concurrently removed snapshots is limited by 2 per single host and 4 in total for the entire cluster. Make sure you're setting the values correctly and restart the backup service after that.

[DonZoomik]

MaxConcurrentDeletingSnapshotsForCluster is set to 8 and MaxConcurrentDeletingSnapshotsForHost to 4 but I only still see concurrent retention policy tasks in replication jobs, even after restart backup service again. Replica VMs are quite evenly spread over hosts as well. It's a bit hard to keep track but backup job snapshot removals seem to have increased but no dice for replication...

[foggy]

Please open a support case for a closer look. Thanks!

[DonZoomik]

#04490270

[DonZoomik]

I had to cancel the original ticket for reasons but I got back to it a bit later.
Support gave me another registry value that allows for more parallel tasks however I've been instructed to not share this publicly. #04548985 for reference if anyone needs it.

[foggy]

Thanks for the update and reference - much appreciated!