On replication stepdown we clear the stop the async updates to the oplogTruncateAfterPoint and then null the value here. On shutdown while primary, we clear the oplogTruncateAfterPoint here. However, we do not stop the async updates to the oplogTruncateAfterPoint on primary shutdown, so the async thread can set the oplogTruncateAfterPoint again before shutdown completes.
- is related to
-
SERVER-47582 Stepdown in drain mode can trigger deadlock due to clearing OplogTruncateAfterPoint
- Closed
-
SERVER-46992 Create JS test for v4.4-v4.2 oplogTruncateAfterPoint compatibility across v4.4 shutdown while primary
- Closed