ray.rllib.evaluation.worker_set.WorkerSet.fetch_ready_async_reqs#

WorkerSet.fetch_ready_async_reqs(*, timeout_seconds: float | None = 0.0, return_obj_refs: bool = False, mark_healthy: bool = True) → List[Tuple[int, T]][source]#

Get esults from outstanding asynchronous requests that are ready.

Parameters:

  • timeout_seconds – Time to wait for results. Default is 0, meaning those requests that are already ready.

  • return_obj_refs – Whether to return ObjectRef instead of actual results.

  • mark_healthy – Whether to mark all those workers healthy again that are currently marked unhealthy AND that returned results from the remote call (within the given timeout_seconds). Note that workers are NOT set unhealthy, if they simply time out (only if they return a RayActorError). Also not that this setting is ignored if healthy_only=True was set in the preceding self.foreach_worker_asyn() call, b/c the mark_healthy setting only affects workers that are currently tagged as unhealthy.

Returns:

A list of results successfully returned from outstanding remote calls, paired with the indices of the callee workers.