How are atomic operations on Resume(Job) and Pause(Job) guaranteed in StdScheduler?
miJasonK opened this issue · 1 comments
miJasonK commented
In the StdScheduler object, if the Remove Job succeeds but the Push Job fails, the task will be lost. How should this scenario be supported?
reugn commented
@miJasonK, this is only relevant for custom JobQueue implementations, as no errors can occur using the standard in-memory queue. And it's a broader question - what happens if the job fails to be rescheduled? I would leave error handling to the implementation, including the application of a retry mechanism and the use of a fallback queue if the backend storage is unavailable.