Steps to reproduce:
Issue is quite hard to reproduce since "CleanupSnapshotInternal" stage takes about one second to complete, but essentially it requires for a small network outage during this stage for the problem to manifest.
1. Force snapshot for Agent.
2. Disable network adapter on the Agent during "CleanupSnapshotInternal" stage.
3. Enable network adapter again in a few seconds.
Actual result:
Transfer failed because of absence of retries on CleanupSnapshotInternal stage
ERROR 2017-05-10T00:07:48 [1674] - Replay.Core.Implementation.Transfer.TransferJob (Transfer ID '29280e83-f7a6-449d-82ab-5b5b4f425387' of volumes \\?\Volume{e0771ba2-e3b7-4827-9e8b-9862ac7dbf8c}\,\\?\Volume{7ce451c9-990c-11e3-93ee-6cae8b3fede3}\,\\.\HarddiskVolume1 from 'SAGSGSR003' Job Id = 29280e83-f7a6-449d-82ab-5b5b4f425387:TransferJob)
Error trying to delete snapshot db757646-f241-415f-bf03-fa33ddedb299 on 'SAGSGSR003'
Expected result:
In case of failed attempt to clean up snapshots, one more attempt to do so should be performed.