In the logical session cache sharding suite (1sec refresh) during the CleanEveryN hook, a BatchWriteOp is destroyed before all of the queued writes are completed. This is during shut down, so it's likely that another thread prematurely destroyed it.
Investigation to follow.
- depends on
-
SERVER-37430 Destroy sharding task executors and AsyncRequestSenders after the PeriodicRunner is destroyed in mongod shutdown
- Closed
- is depended on by
-
SERVER-37330 Add sharded passthrough suites to detect errors in LogicalSessionCache
- Closed