An explanation of the occurrence is in the linked BF.
One solution would be to trigger migration recovery (similar to the recovery on step up) if an error is caught at this stage in the MigrationDestinationManager instead of simply clearing the MigrationDestinationManager's state.
- is related to
-
SERVER-92236 Chunk migrations should use short lived cancellation sources
- Backlog
-
SERVER-66477 Deadlock during stepup when there is a prepared transaction and migration recipient recovery needs to be run
- Closed