apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-12-24 13:28:13 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Feature] EngineWorkerQueue anonymous port (#4597)

Browse Source 
* EngineWorkerQueue 支持匿名端口设置

* EngineWorkerQueue 支持匿名端口设置

* EngineWorkerQueue 支持匿名端口设置

* EngineWorkerQueue 支持匿名端口设置

* EngineWorkerQueue 支持匿名端口设置
This commit is contained in:
Daci
2025-10-28 10:22:37 +08:00
committed by GitHub
parent 7681375a19
commit 6426414a0f
5 changed files with 46 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

12
fastdeploy/engine/common_engine.py
View File

@@ -268,6 +268,16 @@ class EngineService:
num_client=self.cfg.parallel_config.tensor_parallel_size,
local_data_parallel_size=self.cfg.parallel_config.data_parallel_size,
)
# Dynamically updates the port value if an anonymous port is used
self.cfg.parallel_config.engine_worker_queue_port[self.cfg.parallel_config.local_data_parallel_id] = str(
self.engine_worker_queue_server.get_server_port()
)
address = (
self.cfg.master_ip,
int(
self.cfg.parallel_config.engine_worker_queue_port[self.cfg.parallel_config.local_data_parallel_id]
),
)
if self.cfg.cache_config.enable_prefix_caching or self.cfg.scheduler_config.splitwise_role != "mixed":
self.cache_task_queue = EngineCacheQueue(
@@ -281,6 +291,8 @@ class EngineService:
client_id=-1,
local_data_parallel_size=self.cfg.parallel_config.data_parallel_size,
)
self.cfg.cache_config.cache_queue_port = self.cache_task_queue.get_server_port()
self.llm_logger.info(
f"local {min(self.cfg.worker_num_per_node * self.cfg.node_rank + self.cfg.parallel_config.local_data_parallel_id,self.cfg.parallel_config.data_parallel_size - 1)}"
)