hpcflow.sdk.monitoring.active_runs.execution_plans#

hpcflow.sdk.monitoring.active_runs.execution_plans(job)#

Follow generated loop order rather than numeric run-ID order.

Array items execute one column in action order. Serial shell jobs execute block/element/action; combined Python jobs execute block/action/element. Missing -1 slots are omitted; ambiguous repeated IDs disable this optimization.

Parameters:

job (Jobscript)

Return type:

dict[int, tuple[int, …]]