Better PD initialization (#5751)

python/sglang/srt/disaggregation/mini_lb.py

@@ -3,10 +3,12 @@ Minimal HTTP load balancer for prefill and decode servers for testing.
 """
 
 import asyncio
+import dataclasses
+import logging
 import random
 import urllib
 from itertools import chain
-from typing import List
+from typing import List, Optional
 
 import aiohttp
 import orjson
@@ -14,11 +16,32 @@ import uvicorn
 from fastapi import FastAPI, HTTPException
 from fastapi.responses import ORJSONResponse, Response, StreamingResponse
 
+from sglang.srt.disaggregation.utils import PDRegistryRequest
 
+
+def setup_logger():
+    logger = logging.getLogger("pdlb")
+    logger.setLevel(logging.INFO)
+
+    formatter = logging.Formatter(
+        "[PDLB (Python)] %(asctime)s - %(levelname)s - %(message)s",
+        datefmt="%Y-%m-%d %H:%M:%S",
+    )
+
+    handler = logging.StreamHandler()
+    handler.setFormatter(formatter)
+    logger.addHandler(handler)
+
+    return logger
+
+
+logger = setup_logger()
+
+
+@dataclasses.dataclass
 class PrefillConfig:
-    def __init__(self, url: str, bootstrap_port: int):
+    url: str
-        self.url = url
+    bootstrap_port: Optional[int] = None
-        self.bootstrap_port = bootstrap_port
 
 
 class MiniLoadBalancer:
@@ -28,6 +51,10 @@ class MiniLoadBalancer:
         self.decode_servers = decode_servers
 
     def select_pair(self):
+        # TODO: return some message instead of panic
+        assert len(self.prefill_configs) > 0, "No prefill servers available"
+        assert len(self.decode_servers) > 0, "No decode servers available"
+
         prefill_config = random.choice(self.prefill_configs)
         decode_server = random.choice(self.decode_servers)
         return prefill_config.url, prefill_config.bootstrap_port, decode_server
@@ -47,7 +74,7 @@ class MiniLoadBalancer:
                 session.post(f"{decode_server}/{endpoint}", json=modified_request),
             ]
             # Wait for both responses to complete. Prefill should end first.
-            prefill_response, decode_response = await asyncio.gather(*tasks)
+            _, decode_response = await asyncio.gather(*tasks)
 
             return ORJSONResponse(
                 content=await decode_response.json(),
@@ -268,6 +295,32 @@ async def get_models():
             raise HTTPException(status_code=500, detail=str(e))
 
 
+@app.post("/register")
+async def register(obj: PDRegistryRequest):
+    if obj.mode == "prefill":
+        load_balancer.prefill_configs.append(
+            PrefillConfig(obj.registry_url, obj.bootstrap_port)
+        )
+        logger.info(
+            f"Registered prefill server: {obj.registry_url} with bootstrap port: {obj.bootstrap_port}"
+        )
+    elif obj.mode == "decode":
+        load_balancer.decode_servers.append(obj.registry_url)
+        logger.info(f"Registered decode server: {obj.registry_url}")
+    else:
+        raise HTTPException(
+            status_code=400,
+            detail="Invalid mode. Must be either PREFILL or DECODE.",
+        )
+
+    logger.info(
+        f"#Prefill servers: {len(load_balancer.prefill_configs)}, "
+        f"#Decode servers: {len(load_balancer.decode_servers)}"
+    )
+
+    return Response(status_code=200)
+
+
 def run(prefill_configs, decode_addrs, host, port):
     global load_balancer
     load_balancer = MiniLoadBalancer(prefill_configs, decode_addrs)
@@ -279,15 +332,16 @@ if __name__ == "__main__":
 
     parser = argparse.ArgumentParser(description="Mini Load Balancer Server")
     parser.add_argument(
-        "--prefill", required=True, help="Comma-separated URLs for prefill servers"
+        "--prefill", type=str, default=[], nargs="+", help="URLs for prefill servers"
+    )
+    parser.add_argument(
+        "--decode", type=str, default=[], nargs="+", help="URLs for decode servers"
     )
     parser.add_argument(
         "--prefill-bootstrap-ports",
-        help="Comma-separated bootstrap ports for prefill servers",
+        type=int,
-        default="8998",
+        nargs="+",
-    )
+        help="Bootstrap ports for prefill servers",
-    parser.add_argument(
-        "--decode", required=True, help="Comma-separated URLs for decode servers"
     )
     parser.add_argument(
         "--host", default="0.0.0.0", help="Host to bind the server (default: 0.0.0.0)"
@@ -297,22 +351,19 @@ if __name__ == "__main__":
     )
     args = parser.parse_args()
 
-    prefill_urls = args.prefill.split(",")
+    bootstrap_ports = args.prefill_bootstrap_ports
-    bootstrap_ports = [int(p) for p in args.prefill_bootstrap_ports.split(",")]
+    if bootstrap_ports is None:
-
+        bootstrap_ports = [None] * len(args.prefill)
-    if len(bootstrap_ports) == 1:
+    elif len(bootstrap_ports) == 1:
-        bootstrap_ports = bootstrap_ports * len(prefill_urls)
+        bootstrap_ports = bootstrap_ports * len(args.prefill)
     else:
-        if len(bootstrap_ports) != len(prefill_urls):
+        if len(bootstrap_ports) != len(args.prefill):
             raise ValueError(
                 "Number of prefill URLs must match number of bootstrap ports"
             )
-            exit(1)
 
-    prefill_configs = []
+    prefill_configs = [
-    for url, port in zip(prefill_urls, bootstrap_ports):
+        PrefillConfig(url, port) for url, port in zip(args.prefill, bootstrap_ports)
-        prefill_configs.append(PrefillConfig(url, port))
+    ]
 
-    decode_addrs = args.decode.split(",")
+    run(prefill_configs, args.decode, args.host, args.port)
-
-    run(prefill_configs, decode_addrs, args.host, args.port)

python/sglang/srt/disaggregation/utils.py

@@ -1,13 +1,18 @@
 from __future__ import annotations
 
+import dataclasses
+import warnings
 from collections import deque
 from enum import Enum
-from typing import List
+from typing import List, Optional
 
 import numpy as np
+import requests
 import torch
 import torch.distributed as dist
 
+from sglang.srt.utils import get_ip
+
 
 class DisaggregationMode(Enum):
     NULL = "null"
@@ -119,3 +124,41 @@ def kv_to_page_indices(kv_indices: np.ndarray, page_size: int):
 def kv_to_page_num(num_kv_indices: int, page_size: int):
     # ceil(num_kv_indices / page_size)
     return (num_kv_indices + page_size - 1) // page_size
+
+
+@dataclasses.dataclass
+class PDRegistryRequest:
+    """A request to register a machine itself to the LB."""
+
+    mode: str
+    registry_url: str
+    bootstrap_port: Optional[int] = None
+
+    def __post_init__(self):
+        if self.mode == "prefill" and self.bootstrap_port is None:
+            raise ValueError("Bootstrap port must be set in PREFILL mode.")
+        elif self.mode == "decode" and self.bootstrap_port is not None:
+            raise ValueError("Bootstrap port must not be set in DECODE mode.")
+        elif self.mode not in ["prefill", "decode"]:
+            raise ValueError(
+                f"Invalid mode: {self.mode}. Must be 'prefill' or 'decode'."
+            )
+
+
+def register_disaggregation_server(
+    mode: str, server_port: int, bootstrap_port: int, pdlb_url: str
+):
+    boostrap_port = bootstrap_port if mode == "prefill" else None
+    registry_request = PDRegistryRequest(
+        mode=mode,
+        registry_url=f"http://{get_ip()}:{server_port}",
+        bootstrap_port=boostrap_port,
+    )
+    res = requests.post(
+        f"{pdlb_url}/register",
+        json=dataclasses.asdict(registry_request),
+    )
+    if res.status_code != 200:
+        warnings.warn(
+            f"Failed to register disaggregation server: {res.status_code} {res.text}"
+        )

python/sglang/srt/entrypoints/http_server.py

@@ -42,7 +42,10 @@ from fastapi import FastAPI, File, Form, Request, UploadFile
 from fastapi.middleware.cors import CORSMiddleware
 from fastapi.responses import ORJSONResponse, Response, StreamingResponse
 
-from sglang.srt.disaggregation.utils import FakeBootstrapHost
+from sglang.srt.disaggregation.utils import (
+    FakeBootstrapHost,
+    register_disaggregation_server,
+)
 from sglang.srt.entrypoints.engine import _launch_subprocesses
 from sglang.srt.function_call_parser import FunctionCallParser
 from sglang.srt.managers.io_struct import (
@@ -871,5 +874,13 @@ def _wait_and_warmup(
     if server_args.debug_tensor_dump_input_file:
         kill_process_tree(os.getpid())
 
+    if server_args.pdlb_url is not None:
+        register_disaggregation_server(
+            server_args.disaggregation_mode,
+            server_args.port,
+            server_args.disaggregation_bootstrap_port,
+            server_args.pdlb_url,
+        )
+
     if launch_callback is not None:
         launch_callback()

python/sglang/srt/managers/scheduler.py

@@ -925,6 +925,10 @@ class Scheduler(
                 )
                 custom_logit_processor = None
 
+            if recv_req.bootstrap_port is None:
+                # Use default bootstrap port
+                recv_req.bootstrap_port = self.server_args.disaggregation_bootstrap_port
+
             req = Req(
                 recv_req.rid,
                 recv_req.input_text,

python/sglang/srt/server_args.py

@@ -198,6 +198,7 @@ class ServerArgs:
     disaggregation_bootstrap_port: int = 8998
     disaggregation_transfer_backend: str = "mooncake"
     disaggregation_ib_device: Optional[str] = None
+    pdlb_url: Optional[str] = None
 
     def __post_init__(self):
         # Expert parallelism
@@ -1254,6 +1255,12 @@ class ServerArgs:
             "or multiple comma-separated devices (e.g., --disaggregation-ib-device mlx5_0,mlx5_1). "
             "Default is None, which triggers automatic device detection when mooncake backend is enabled.",
         )
+        parser.add_argument(
+            "--pdlb-url",
+            type=str,
+            default=None,
+            help="The URL of the PD disaggregation load balancer. If set, the prefill/decode server will register with the load balancer.",
+        )
 
     @classmethod
     def from_cli_args(cls, args: argparse.Namespace):