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
Workers are not currently aware of which schedule generated a flow run. After we start tracking this with the CreatedBy column, the worker can find the schedule using that field.
When a deployment schedule has max_active_runs, the worker should use a global concurrency limit whose slots are equal to max_active_runs
This will require the client-side interface to global concurrency limits to support a timeout option
The worker could create this limit but will need to fail gracefully if another worker tries to create the same limit at the same time (i.e. there should be a unique constraint)
We could also create the limit when editing or creating the deployment schedule, in which case the worker should fail gracefully if no limit exists (it shouldn't run the run and should probably leave it scheduled)
If the worker can't acquire a slot within a timeout period, it will mark the run as Cancelled with the name Skipped
The text was updated successfully, but these errors were encountered:
CreatedBy
column, the worker can find the schedule using that field.max_active_runs
, the worker should use a global concurrency limit whose slots are equal tomax_active_runs
The text was updated successfully, but these errors were encountered: