Sometimes the publication worker that receive the revokation task is so fast that it calls the orchestrator back before the internal state of the api core is updated.
A simple but not robust trick is to add a time.sleep() to the revoke method of the publication worker however it would be better to do as following, in the revoke method of the api:
generate an uuid
update the publication task with the revoke_task_id = uuid
launch the revoke task with id = uuid
rollback the publication task in case of failure (?)
Sometimes the publication worker that receive the revokation task is so fast that it calls the orchestrator back before the internal state of the api core is updated.
A simple but not robust trick is to add a time.sleep() to the revoke method of the publication worker however it would be better to do as following, in the revoke method of the api:
No aliasing is possible.