克隆Conda环境后执行ray.init()出现PermissionError问题求助

问题描述 #

将Conda环境克隆到新电脑后运行Ray项目，执行ray.init()时遭遇权限错误，具体报错如下：

>>> import ray
>>> ray.init()
2023-11-17 17:26:16,023 INFO resource_spec.py:212 -- Starting Ray with 33.89 GiB memory available for workers and up to 16.97 GiB for objects. You can adjust these settings with ray.init(memory=<bytes>, object_store_memory=<bytes>).
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/root/anaconda3/envs/ncd/lib/python3.7/site-packages/ray/worker.py", line 767, in init
    ray_params=ray_params)
  File "/root/anaconda3/envs/ncd/lib/python3.7/site-packages/ray/node.py", line 177, in __init__
    self.start_head_processes()
  File "/root/anaconda3/envs/ncd/lib/python3.7/site-packages/ray/node.py", line 649, in start_head_processes
    self.start_redis()
  File "/root/anaconda3/envs/ncd/lib/python3.7/site-packages/ray/node.py", line 473, in start_redis
    fate_share=self.kernel_fate_share)
  File "/root/anaconda3/envs/ncd/lib/python3.7/site-packages/ray/services.py", line 769, in start_redis
    fate_share=fate_share)
  File "/root/anaconda3/envs/ncd/lib/python3.7/site-packages/ray/services.py", line 943, in _start_redis_instance
    fate_share=fate_share)
  File "/root/anaconda3/envs/ncd/lib/python3.7/site-packages/ray/services.py", line 514, in start_ray_process
    preexec_fn=preexec_fn if sys.platform != "win32" else None)
  File "/root/anaconda3/envs/ncd/lib/python3.7/subprocess.py", line 800, in __init__
    restore_signals, start_new_session)
  File "/root/anaconda3/envs/ncd/lib/python3.7/subprocess.py", line 1551, in _execute_child
    raise child_exception_type(errno_num, err_msg, err_filename)
PermissionError: [Errno 13] Permission denied: '/root/anaconda3/envs/ncd/lib/python3.7/site-packages/ray/core/src/ray/thirdparty/redis/src/redis-server'

已确认环境目录权限为drwxr-xr-x 12 root root 4096 Nov 17 16:53 ncd，但问题仍存在。

解决方案 #

1. 修复Redis可执行文件的执行权限 #

报错核心是redis-server文件缺少执行权限，先检查该文件的权限：

ls -l /root/anaconda3/envs/ncd/lib/python3.7/site-packages/ray/core/src/ray/thirdparty/redis/src/redis-server

如果输出中文件权限不含x（执行权限），执行以下命令添加：

chmod +x /root/anaconda3/envs/ncd/lib/python3.7/site-packages/ray/core/src/ray/thirdparty/redis/src/redis-server

2. 调整环境目录的所有权（若当前非root用户运行） #

环境位于/root目录下，普通用户默认无读写执行权限。将目录所有权切换为当前用户：

chown -R 你的用户名:你的用户名 /root/anaconda3/envs/ncd

替换你的用户名为实际登录的用户名，之后重新激活环境运行Ray。

3. 重新安装Ray到克隆环境 #

克隆Conda环境可能导致二进制文件依赖或权限损坏，直接重新安装Ray：

conda activate ncd
pip uninstall -y ray
pip install ray

内容的提问来源于stack exchange，提问作者BobCheng