You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There're schedules and REST API (PUT /api/attempts) that issue new attempts. require> also issues a new attempt but it should be implemented as an alias of REST API.
The text was updated successfully, but these errors were encountered:
Schedule: if number of active attempts reached hard limit, starting a new schedule delays for several minutes (1 minute first, 10 minutes maximum, using exponential back-off is ideal). After the wait, scheduler tries to issue attempts again. I think this is good because users have chance to fix the schedule when scheduling is blocked.
REST API: if number of active attempts reached hard limit, sever returns 429 Too Many Requests or 400 Bad Request.
Soft limit: if number of active attempts reached soft limit, scheduler and REST API put a new attempt to a queue and don't put tasks. Workflow executor puts tasks of the attempts later. Soft limit could be used also for multi-tenant scheduling.
There're schedules and REST API (
PUT /api/attempts
) that issue new attempts.require>
also issues a new attempt but it should be implemented as an alias of REST API.The text was updated successfully, but these errors were encountered: