-
Type:
Task
-
Resolution: Duplicate
-
Priority:
Major - P3
-
None
-
Affects Version/s: None
-
Component/s: None
-
None
-
Storage Execution
-
None
-
None
-
None
-
None
-
None
-
None
-
None
The applier honors rids iff the node holds a groupId, replacing the per-collection catalog check. This is the core of the feature: in-group nodes apply updates and deletes by rid, out-of-group nodes fall back to _id lookup.
Details
- Relax tassert 11454701 (rid presence iff collection option) to a membership-plus-in-scope check.
- In-group: inserts at the replicated rid, updates and deletes located by rid (existing computeEffectiveRid machinery).
- Out-of-group: inserts generate local rids (today's non-rrid behavior), updates and deletes fall back to _id-based lookup.
- Rid honoring is active in all application modes: steady state, initial sync application phase, and startup oplog recovery.
Acceptance criteria
- Unit tests for in-group and out-of-group application of insert/update/delete.
- jstest with a mixed-membership replica set validating both paths.