-
Type: Improvement
-
Status: Resolved
-
Priority: Minor
-
Resolution: Fixed
-
Affects Version/s: 6.0
-
Component/s: Events / Works
-
Tags:
The WorkManagerImpl.awaitCompletion implement rely on a Java Condition lock that works only with a standalone Nuxeo instance.
This method should rely on message queue size so it will work in cluster mode when using a Redis job queue.
- is required by
-
NXP-20519 leaked scheduled works in redis after shutdown
- Resolved
SUCCESS: Integrated in nuxeo-master #1270
NXP-18550: indent (jenkins: 7947879ef4fb05a83f31cf88303291130539d362)NXP-18550: Don't rely on the work queue to list scheduled job, the queue is managed by the TP and there is no state when the work is taken from the queue and not yet run. Use a dedicated scheduled set instead. (jenkins: e320b1ba6e6347c26447b5a65bb4929d37a859fd)NXP-18550: fix manual RedisFeature setup that don't init the workmanager (jenkins: 481215126aedb95d37c10cb160d686088fefac3b)