[feat] Support cross-job actor discovery via explicit namespace

When multiple Ray Jobs share the same Ray cluster, Named Actors are
isolated by namespace. Without an explicit namespace, a TQ Controller
created by one job is invisible to workers in another job.

This commit adds namespace="transfer_queue" to both:
- ray.get_actor() in _init_from_existing()
- TransferQueueController.options() in init()

This ensures that the TQ Controller is always registered and discovered
in the fixed "transfer_queue" namespace, enabling cross-job TQ sharing
(e.g., a teacher server job creates TQ, and a trainer job connects to it).

This change is backward-compatible: single-job usage is unaffected since
the namespace is consistent between creation and discovery.

Author: huniu20

transfer_queue/interface.py

@@ -93,7 +93,7 @@ def _init_from_existing() -> bool:
     global _TQ_CONTROLLER
     try:
         if _TQ_CONTROLLER is None:
-            _TQ_CONTROLLER = ray.get_actor("TransferQueueController")
+            _TQ_CONTROLLER = ray.get_actor("TransferQueueController", namespace="transfer_queue")
 
     except ValueError:
         logger.info("Called _init_from_existing() but TransferQueueController has not been initialized yet.")
@@ -174,7 +174,7 @@ def init(conf: DictConfig | None = None) -> DictConfig | None:
 
     try:
         global _TQ_CONTROLLER
-        _TQ_CONTROLLER = TransferQueueController.options(name="TransferQueueController").remote(  # type: ignore[attr-defined]
+        _TQ_CONTROLLER = TransferQueueController.options(name="TransferQueueController", namespace="transfer_queue").remote(  # type: ignore[attr-defined]
             sampler=sampler, polling_mode=final_conf.controller.polling_mode
         )
         logger.info("TransferQueueController has been created.")