Tasks not delivered to resource-manager are not cleaned up
CLOSED - DUPLICATE
If task was not delivered to resource manager because of redis queue failure, it will forever stay in waiting state.
Steps to reproduce:
- Start API service.
- Trigger a task. Check a record is created in table
core_taskand it is in
- Kill redis. (The best way to imitate the behavior is to disable pestistent storage for redis).
- Start redis.
- Start resource manager.
- Start worker.
- Task created on step 2 is not cleaned up from
core_taskdatabase table and persists in