hpcflow.sdk.monitoring.runs.collect_runs#

hpcflow.sdk.monitoring.runs.collect_runs(workflow, state=None, *, task_idx=None, task_id=None, task_name=None, element_idx=None, iteration_idx=None, action_idx=None, run_id=None, run_idx=None, columns=None)#

Read current run states and return matching attempts in run-ID order.

Include all iterations, submissions, and attempts, including failed attempts subsequently retried successfully. Use a shared store-cache context for bulk retrieval. No scheduler queries, registry updates, or terminal-cache writes are performed. Invalid state names fail before reading the workflow store. Coordinates are ANDed; comma-separated alternatives and inclusive ranges are accepted within each numeric selection. Task indices refer to workflow order, task IDs are stable insertion IDs, and run indices are per-action attempts. An explicit run-ID selection loads only those valid IDs. Array positions are read only when requested in columns (or for direct calls without columns). Duration uses recorded endpoints, or one current instant for running runs.

Parameters:
  • workflow (Workflow)

  • state (EARStatus | str | None)

  • task_idx (IndexFilter)

  • task_id (IndexFilter)

  • task_name (str | Sequence[str] | None)

  • element_idx (IndexFilter)

  • iteration_idx (IndexFilter)

  • action_idx (IndexFilter)

  • run_id (IndexFilter)

  • run_idx (IndexFilter)

  • columns (str | Sequence[str] | None)

Return type:

list[RunRecord]