Closed lackita closed 7 years ago
@lackita Great catch, had no idea apply_aysnc is re-called when retrying a task. Also, apologises for insane delay in getting to this PR, I've feel out of sync maintaining this project big time. With the new 1.0 release, I've ported your work to #48 and will get it released shortly to pypi.
Thanks again for the PR.
This is released.
When .retry is called, the state of the task becomes such that after_return is not called. This is probably correct on some level, since retrying means a new version will be queued. Unfortunately, the requeue doesn't happen because it attempts to reacquire the lock, leaving a lock in redis for a task not in any queue.