Each ReshardingOplogApplier instance (one per donor shard) currently receives its own ThreadPool for actually performing oplog application. Sizing the number of threads to use for resharding would be simpler for end users if there was only one of them, i.e. RecipientStateMachine's task executor.
- depends on
-
SERVER-55384 Move session application for resharding's oplog application into its own class
- Closed
- is duplicated by
-
SERVER-53139 Don't wait for prepared transactions in resharding oplog applier writer thread
- Closed