Skip to content

分布式 worker 端口异常占用,,辛苦看下如何解决 #71

@xuelei123-xgb

Description

@xuelei123-xgb

728604d04ba41d42adb9ce801000000, repr=<verl.single_controller.ray.base.WorkerDict object at 0x7fa4d9b7f590>)
(WorkerDict pid=113813) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(WorkerDict pid=113813) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(WorkerDict pid=113813) File "/data/team/xuelei12/WebAgent/RL_Factory/RL-Factory/verl/single_controller/ray/base.py", line 723, in init
(WorkerDict pid=113813) self.worker_dict[key] = user_defined_cls(*init_args_dict[key].get("args", ()), **init_args_dict[key].get("kwargs", {}))
(WorkerDict pid=113813) ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
(WorkerDict pid=113813) File "/data/team/xuelei12/WebAgent/RL_Factory/RL-Factory/verl/workers/fsdp_workers.py", line 112, in init
(WorkerDict pid=113813) torch.distributed.init_process_group(backend="cpu:gloo,cuda:nccl" if is_cuda_available else "cpu:gloo,npu:hccl", rank=rank, world_size=world_size)
(WorkerDict pid=113813) File "/root/miniconda3/envs/rl_facory_py311/lib/python3.11/site-packages/torch/distributed/c10d_logger.py", line 81, in wrapper
(WorkerDict pid=113813) return func(*args, **kwargs)
(WorkerDict pid=113813) ^^^^^^^^^^^^^^^^^^^^^
(WorkerDict pid=113813) File "/root/miniconda3/envs/rl_facory_py311/lib/python3.11/site-packages/torch/distributed/c10d_logger.py", line 95, in wrapper
(WorkerDict pid=113813) func_return = func(*args, **kwargs)
(WorkerDict pid=113813) ^^^^^^^^^^^^^^^^^^^^^
(WorkerDict pid=113813) File "/root/miniconda3/envs/rl_facory_py311/lib/python3.11/site-packages/torch/distributed/distributed_c10d.py", line 1714, in init_process_group
(WorkerDict pid=113813) store, rank, world_size = next(rendezvous_iterator)
(WorkerDict pid=113813) ^^^^^^^^^^^^^^^^^^^^^^^^^
(WorkerDict pid=113813) File "/root/miniconda3/envs/rl_facory_py311/lib/python3.11/site-packages/torch/distributed/rendezvous.py", line 274, in _env_rendezvous_handler
(WorkerDict pid=113813) store = _create_c10d_store(
(WorkerDict pid=113813) ^^^^^^^^^^^^^^^^^^^
(WorkerDict pid=113813) File "/root/miniconda3/envs/rl_facory_py311/lib/python3.11/site-packages/torch/distributed/rendezvous.py", line 194, in _create_c10d_store
(WorkerDict pid=113813) return TCPStore(
(WorkerDict pid=113813) ^^^^^^^^^
(WorkerDict pid=113813) RuntimeError: The server socket has failed to listen on any local network address. port: 60107, useIpv6: 0, code: -98, name: EADDRINUSE, message: address already in use

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions