matrice-compute 0.1.21__tar.gz → 0.1.22__tar.gz

{matrice_compute-0.1.21 → matrice_compute-0.1.22}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: matrice_compute
-Version: 0.1.21
+Version: 0.1.22
 Summary: Common server utilities for Matrice.ai services
 Author-email: "Matrice.ai" <dipendra@matrice.ai>
 License-Expression: MIT

{matrice_compute-0.1.21 → matrice_compute-0.1.22}/matrice_compute.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: matrice_compute
-Version: 0.1.21
+Version: 0.1.22
 Summary: Common server utilities for Matrice.ai services
 Author-email: "Matrice.ai" <dipendra@matrice.ai>
 License-Expression: MIT

{matrice_compute-0.1.21 → matrice_compute-0.1.22}/src/matrice_compute/scaling.py

@@ -677,7 +677,7 @@ class Scaling:
 
     @log_errors(log_error=True)
     def assign_jobs(self, is_gpu):
-        """Assign jobs to the instance using Kafka (with REST fallback).
+        """Assign jobs to the instance using REST API.
 
         Args:
             is_gpu: Boolean or any value indicating if this is a GPU instance.
@@ -690,30 +690,41 @@ class Scaling:
         is_gpu_bool = bool(is_gpu)
         logging.info("Assigning jobs for instance %s (GPU: %s)", self.instance_id, is_gpu_bool)
 
-        payload = {
-            "instanceID": self.instance_id,
-            "isGPUInstance": is_gpu_bool,
-        }
-
-        # Define REST fallback function
-        def rest_fallback():
-            is_gpu_str = str(is_gpu_bool).lower()
-            path = f"/v1/actions/assign_jobs/{is_gpu_str}/{self.instance_id}"
-            resp = self.rpc.get(path=path)
-            return self.handle_response(
-                resp,
-                "Pinged successfully",
-                "Could not ping the scaling jobs",
-            )
-
-        # Use hybrid approach: Kafka first, REST fallback, cache if both fail
-        return self._hybrid_request(
-            api="assign_jobs",
-            payload=payload,
-            request_topic=self.kafka_config["action_request_topic"] if self.enable_kafka else None,
-            response_topic=self.kafka_config["action_response_topic"] if self.enable_kafka else None,
-            rest_fallback_func=rest_fallback
+        # Use REST API directly
+        is_gpu_str = str(is_gpu_bool).lower()
+        path = f"/v1/actions/assign_jobs/{is_gpu_str}/{self.instance_id}"
+        resp = self.rpc.get(path=path)
+        return self.handle_response(
+            resp,
+            "Pinged successfully",
+            "Could not ping the scaling jobs",
         )
+
+        # # Kafka approach (commented out - using REST only)
+        # payload = {
+        #     "instanceID": self.instance_id,
+        #     "isGPUInstance": is_gpu_bool,
+        # }
+        # 
+        # # Define REST fallback function
+        # def rest_fallback():
+        #     is_gpu_str = str(is_gpu_bool).lower()
+        #     path = f"/v1/actions/assign_jobs/{is_gpu_str}/{self.instance_id}"
+        #     resp = self.rpc.get(path=path)
+        #     return self.handle_response(
+        #         resp,
+        #         "Pinged successfully",
+        #         "Could not ping the scaling jobs",
+        #     )
+        # 
+        # # Use hybrid approach: Kafka first, REST fallback, cache if both fail
+        # return self._hybrid_request(
+        #     api="assign_jobs",
+        #     payload=payload,
+        #     request_topic=self.kafka_config["action_request_topic"] if self.enable_kafka else None,
+        #     response_topic=self.kafka_config["action_response_topic"] if self.enable_kafka else None,
+        #     rest_fallback_func=rest_fallback
+        # )
           
 
     @log_errors(log_error=True)