-
Type:
Task
-
Resolution: Won't Fix
-
Priority:
Major - P3
-
None
-
Affects Version/s: None
-
Component/s: None
-
None
-
Catalog and Routing
-
v8.0
-
CAR Team 2026-08-31
-
0
-
None
-
None
-
None
-
None
-
None
-
None
-
None
mongos_writes_wait_for_write_concern_timeseries.js still hits Evergreen idle timeout after SERVER-127770. The helper sends wtimeout: 1000, but sharded DDL forwarded via appendMajorityWriteConcern waits on majority with a 60s floor (enableSharding, setFCV, create, dropIndexes). One secondary is down and another is {{fsyncLock}}ed, so those waits run to completion.
As part of this ticket, we need to cap how many long waits stack in one file so that the Evergreen idle timeout is not hit. This change is isolated to v8.0 since SERVER-102770 / SERVER-103553 are not backported to v8.0, therefore v8.2+ are not affected.
- is related to
-
SERVER-102770 Refactor code using internal sharding write concern timeouts
-
- Closed
-
-
SERVER-103553 Improve test runtime of jstests/sharding/mongos_writes_wait_for_write_concern_timeseries.js
-
- Closed
-
-
SERVER-127770 Reduce the window of running commands in mongos_writes_wait_for_write_concern_timeseries to 10%
-
- Closed
-