[Data] "Actor ... has constructor arguments in the object store" warning with minimal Ray Data use #57733
Description
What happened + What you expected to happen
I was running a regular batch inference pipeline, and saw this error. This confused me, because I'm not using any large actor arguments.
Running 0: 0.00 row [00:00, ? row/s]{"asctime":"2025-10-15 07:54:31,170","levelname":"E","message":"Actor with class name: 'MapWorker(MapBatches(EmbedPatches))' and ID: '9e175f8de7f8d1f2271c24e308000000' has constructor arguments in the object store and max_restarts > 0. If the arguments in the object store go out of scope or are lost, the actor restart will fail. See #53727 for more details.","filename":"core_worker.cc","lineno":2170}
Versions / Dependencies
Reproduction script
>>> class Actor:
... def __call__(self, batch):
... return batch
...
>>> ray.data.range(1).map_batches(Actor).materialize()2025-10-15 10:08:45,989 INFO logging.py:293 -- Registered dataset logger for dataset dataset_6_0
2025-10-15 10:08:45,993 INFO streaming_executor.py:159 -- Starting execution of Dataset dataset_6_0. Full logs are in /tmp/ray/session_2025-10-15_10-08-22_304424_12580/logs/ray-data
**2025-10-15 10:08:45,994 INFO streaming_executor.py:160 -- Execution plan of Dataset dataset_6_0: InputDataBuffer[Input] -> TaskPoolMapOperator[ReadRange] -> ActorPoolMapOperator[MapBatches(Actor)]**
THIS IS THE ISSUE -> [2025-10-15 10:08:46,014 E 12580 63854] core_worker.cc:2162: Actor with class name: 'MapWorker(MapBatches(Actor))' and ID: 'f317cd9c02fcc41ff203c17201000000' has constructor arguments in the object store and max_restarts > 0. If the arguments in the object store go out of scope or are lost, the actor restart will fail. See https://github.com/ray-project/ray/issues/53727 for more details.
2025-10-15 10:08:46,319 INFO streaming_executor.py:279 -- ✔️ Dataset dataset_6_0 execution finished in 0.33 seconds
✔️ Dataset dataset_6_0 execution finished in 0.33 seconds: 100%|██████████████████████████████████████████████████████████████████| 1.00/1.00 [00:00<00:00, 3.06 row/s]
- ReadRange->SplitBlocks(20): Tasks: 0; Actors: 0; Queued blocks: 0; Resources: 0.0 CPU, 0.0B object store: 100%|██████████████████| 1.00/1.00 [00:00<00:00, 3.26 row/s]
- MapBatches(Actor): Tasks: 0; Actors: 0; Queued blocks: 0; Resources: 1.0 CPU, 8.0B object store; [0/1 objects local]: : 1.00 row [00:00, 3.25 row/s]
MaterializedDataset(num_blocks=1, num_rows=1, schema={id: int64})Issue Severity
None