[core][distributed] fix ray worker rank assignment (vllm-project#6235)

vllm/executor/ray_gpu_executor.py

@@ -134,11 +134,32 @@ def _init_workers_ray(self, placement_group: "PlacementGroup",
         worker_node_and_gpu_ids = self._run_workers("get_node_and_gpu_ids",
                                                     use_dummy_driver=True)
 
-        node_workers = defaultdict(list)
-        node_gpus = defaultdict(list)
-
-        for i, (node_id, gpu_ids) in enumerate(worker_node_and_gpu_ids):
-            node_workers[node_id].append(i)
+        # the order in `worker_node_and_gpu_ids` does not necessarily match
+        # the machine boundaries. We need to make sure that workers in the
+        # same node are assigned consecutive ranks.
+        # examples:
+        # [('852a09a13c7503ef126d7c828454c741494b1be33a8627a5206604d9', [0]), ('dfaad7adfdae57a694cc74490db45bd112c9f31243523e43ddc2e7f0', [0]), ('dfaad7adfdae57a694cc74490db45bd112c9f31243523e43ddc2e7f0', [1]), ('dfaad7adfdae57a694cc74490db45bd112c9f31243523e43ddc2e7f0', [2]), ('dfaad7adfdae57a694cc74490db45bd112c9f31243523e43ddc2e7f0', [3]), ('852a09a13c7503ef126d7c828454c741494b1be33a8627a5206604d9', [1]), ('852a09a13c7503ef126d7c828454c741494b1be33a8627a5206604d9', [2]), ('852a09a13c7503ef126d7c828454c741494b1be33a8627a5206604d9', [3])] # noqa
+
+        # initialize worker ranks with -1 (unassigned)
+        worker_ranks = [-1 for x in worker_node_and_gpu_ids]
+        current_rank = 0
+        while -1 in worker_ranks:
+            # whenever we find an unassigned worker, find the node
+            index = worker_ranks.index(-1)
+            current_node_id = worker_node_and_gpu_ids[index][0]
+            # assign ranks to all workers in the same node
+            for i, (node_id, _) in enumerate(worker_node_and_gpu_ids):
+                if node_id == current_node_id:
+                    worker_ranks[i] = current_rank
+                    current_rank += 1
+        # with the above example, worker_ranks will be [0, 4, 5, 6, 7, 1, 2, 3]
+
+        node_workers = defaultdict(list)  # node id -> list of worker ranks
+        node_gpus = defaultdict(list)  # node id -> list of gpu ids
+
+        for worker_rank, (node_id, gpu_ids) in zip(worker_ranks,
+                                                   worker_node_and_gpu_ids):
+            node_workers[node_id].append(worker_rank)
             # `gpu_ids` can be a list of strings or integers.
             # convert them to integers for consistency.
             # NOTE: gpu_ids can be larger than 9 (e.g. 16 GPUs),
@@ -184,7 +205,8 @@ def _init_workers_ray(self, placement_group: "PlacementGroup",
                 local_rank=node_workers[node_id].index(rank),
                 rank=rank,
                 distributed_init_method=distributed_init_method,
-            ) for rank, (node_id, _) in enumerate(worker_node_and_gpu_ids)
+            ) for rank, (node_id,
+                         _) in zip(worker_ranks, worker_node_and_gpu_ids)
         ]
         self._run_workers("init_worker", all_kwargs=init_worker_all_kwargs)