[Bugfix][Nixl] Fix DP Metadata Handshake (#19008)

Signed-off-by: rshaw@neuralmagic.com <robertgshaw2@gmail.com>

vllm/distributed/kv_transfer/kv_connector/v1/nixl_connector.py

@@ -19,7 +19,7 @@ from vllm.distributed.kv_transfer.kv_connector.v1.base import (
     KVConnectorBase_V1, KVConnectorMetadata, KVConnectorRole)
 from vllm.distributed.parallel_state import (
     get_tensor_model_parallel_rank, get_tensor_model_parallel_world_size,
-    get_tp_group, get_world_group)
+    get_tp_group)
 from vllm.logger import init_logger
 from vllm.utils import make_zmq_path, make_zmq_socket, round_down
 from vllm.v1.core.sched.output import SchedulerOutput
@@ -172,6 +172,11 @@ class NixlConnectorScheduler:
         self.vllm_config = vllm_config
         self.block_size = vllm_config.cache_config.block_size
         self.engine_id = engine_id
+        self.side_channel_host = envs.VLLM_NIXL_SIDE_CHANNEL_HOST
+        self.side_channel_port = (
+            envs.VLLM_NIXL_SIDE_CHANNEL_PORT +
+            vllm_config.parallel_config.data_parallel_rank_local *
+            vllm_config.parallel_config.tensor_parallel_size)
         logger.info("Initializing NIXL Scheduler %s", engine_id)
 
         # Requests that need to start recv.
@@ -310,8 +315,8 @@ class NixlConnectorScheduler:
             do_remote_decode=False,
             remote_block_ids=computed_block_ids,
             remote_engine_id=self.engine_id,
-            remote_host=envs.VLLM_NIXL_SIDE_CHANNEL_HOST,
+            remote_host=self.side_channel_host,
-            remote_port=envs.VLLM_NIXL_SIDE_CHANNEL_PORT,
+            remote_port=self.side_channel_port,
         )
 
 
@@ -330,11 +335,19 @@ class NixlConnectorWorker:
         # Map of engine_id -> agent_name.
         self._remote_agents: dict[str, str] = {}
 
+        # NIXL handshake port.
+        # NOTE(rob): Within a DP group, each DP rank gets its own
+        # base port (which is sent in the KVTransferParams).
+        # Each TP rank listens/queries on the base_port + tp_rank.
+        self.side_channel_port = (
+            envs.VLLM_NIXL_SIDE_CHANNEL_PORT +
+            vllm_config.parallel_config.data_parallel_rank_local *
+            vllm_config.parallel_config.tensor_parallel_size)
+
         # Metadata.
         self.engine_id = engine_id
-        self.rank = get_tensor_model_parallel_rank()
+        self.tp_rank = get_tensor_model_parallel_rank()
         self.world_size = get_tensor_model_parallel_world_size()
-        self.world_rank = get_world_group().rank_in_group
         self.tp_group = get_tp_group()
 
         # KV Caches and nixl tracking data.
@@ -383,16 +396,11 @@ class NixlConnectorWorker:
 
     @staticmethod
     def _nixl_handshake_listener(metadata: NixlAgentMetadata,
-                                 ready_event: threading.Event,
+                                 ready_event: threading.Event, base_port: int,
-                                 world_rank: int):
+                                 tp_rank: int):
         """Background thread for getting new NIXL handshakes."""
         # NOTE(rob): this is a simple implementation. We will move
-        # to a better approach like an ETCD server in the future.
+        # to a better approach via HTTP endpoint soon.
-
-        # NOTE(rob): to support heterogeneous TP, we will have to
-        # move this into the scheduler rather than worker, since
-        # each rank needs the metadata of all other ranks (whereas
-        # in this setup, each rank only gets one other rank's meta.
 
         encoder = msgspec.msgpack.Encoder()
         encoded_data = encoder.encode(metadata)
@@ -402,11 +410,7 @@ class NixlConnectorWorker:
 
         # Listen for new requests for metadata.
         host = envs.VLLM_NIXL_SIDE_CHANNEL_HOST
-        # NOTE(rob): we need each rank to have a unique port. This
+        path = make_zmq_path("tcp", host, base_port + tp_rank)
-        # hack to keeps us moving. We will switch when moving to etcd
-        # or where we have a single ZMQ socket in the scheduler.
-        port = envs.VLLM_NIXL_SIDE_CHANNEL_PORT + world_rank
-        path = make_zmq_path("tcp", host, port)
         logger.debug("Starting listening on path: %s", path)
         with zmq_ctx(zmq.ROUTER, path) as sock:
             ready_event.set()
@@ -421,10 +425,10 @@ class NixlConnectorWorker:
         """Do a NIXL handshake with a remote instance."""
 
         start_time = time.perf_counter()
-        # NOTE(rob): we need each rank to have a unique port. This is
+        # NOTE(rob): we need each tp_rank to have a unique port.
-        # a hack to keep us moving. We will switch when moving to etcd
+        # This is a hack to keep us moving. We will switch when
-        # or where we have a single ZMQ socket in the scheduler.
+        # we switch to HTTP-based NIXL metadata exchange.
-        path = make_zmq_path("tcp", host, port + self.world_rank)
+        path = make_zmq_path("tcp", host, port + self.tp_rank)
         logger.debug("Querying metadata on path: %s", path)
         with zmq_ctx(zmq.REQ, path) as sock:
             # Send query for the request.
@@ -532,7 +536,7 @@ class NixlConnectorWorker:
         ready_event = threading.Event()
         self._nixl_handshake_listener_t = threading.Thread(
             target=self._nixl_handshake_listener,
-            args=(metadata, ready_event, self.world_rank),
+            args=(metadata, ready_event, self.side_channel_port, self.tp_rank),
             daemon=True,
             name="nixl_handshake_listener")
         self._nixl_handshake_listener_t.start()
@@ -556,9 +560,9 @@ class NixlConnectorWorker:
                 block_offset = block_id * self.block_len
                 # (addr, len, device id)
                 blocks_data.append(
-                    (base_addr + block_offset, self.block_len, self.rank))
+                    (base_addr + block_offset, self.block_len, self.tp_rank))
-        logger.debug("Created %s blocks for src engine %s and rank %s",
+        logger.debug("Created %s blocks for src engine %s and tp_rank %s",
-                     len(blocks_data), self.engine_id, self.rank)
+                     len(blocks_data), self.engine_id, self.tp_rank)
 
         # Register with NIXL.
         descs = self.nixl_wrapper.get_xfer_descs(blocks_data, "VRAM")
@@ -573,9 +577,9 @@ class NixlConnectorWorker:
                 block_offset = block_id * self.block_len
                 # (addr, len, device id)
                 blocks_data.append(
-                    (base_addr + block_offset, self.block_len, self.rank))
+                    (base_addr + block_offset, self.block_len, self.tp_rank))
-        logger.debug("Created %s blocks for dst engine %s and rank %s",
+        logger.debug("Created %s blocks for dst engine %s and tp_rank %s",
-                     len(blocks_data), engine_id, self.rank)
+                     len(blocks_data), engine_id, self.tp_rank)
 
         # Register with NIXL.
         descs = self.nixl_wrapper.get_xfer_descs(blocks_data, "VRAM")
@@ -600,14 +604,14 @@ class NixlConnectorWorker:
         if len(done_sending) > 0 or len(done_recving) > 0:
             logger.debug(
                 "Rank %s, get_finished: %s requests done sending "
-                "and %s requests done recving", self.rank, len(done_sending),
+                "and %s requests done recving", self.tp_rank,
-                len(done_recving))
+                len(done_sending), len(done_recving))
 
         if self.world_size == 1:
             return done_sending, done_recving
 
         # Rank 0: get finished from all other ranks.
-        if self.rank == 0:
+        if self.tp_rank == 0:
             for req_id in done_sending:
                 self._done_sending_count[req_id] += 1
             for req_id in done_recving: