-
Type: Bug
-
Resolution: Duplicate
-
Priority: Major - P3
-
None
-
Affects Version/s: 3.0.6, 3.1.8
-
Component/s: Replication
-
None
-
ALL
Due to the way sync_tail and bgsync interact while applying and fetching documents it is possible both, during a rollback, write to the minvalid collection with different values.
Since rollback will transition to a non-readable state, like rollback and the applier stops it should be safe but more investigation should be taken.
And we should probably disallow this race all together, by ensuring the batch completes before we enter rollback.
- duplicates
-
SERVER-21988 Rollback does not wait for applier to finish before starting
- Closed
- related to
-
SERVER-20326 Record "apply" batch boundaries during replication
- Closed