In some situation it appears that some node task takes too long for some reasons.
In order to not block the service forever, we might want to cancel a task to allow the task list to continue to be processed.
Cancellation is indeed what we can think of in the first place to solve this kind of problem. But it brings quite some questions:
cancelling a task means we stop the execution of an action in the middle, which could cause inconsistent state of the service, how do we prevent that
In some situation it appears that some node task takes too long for some reasons. In order to not block the service forever, we might want to cancel a task to allow the task list to continue to be processed.
Cancellation is indeed what we can think of in the first place to solve this kind of problem. But it brings quite some questions:
From a code perspective I don't think implementing this is going to be hard. From the top of my head I can see:
/services/{service_guid}/task_list/{task_guid}