使用第三方Git仓库的SageMaker MLOps项目克隆失败求助

问题分析 #

这是SageMaker Studio关联CodePipeline MLOps项目时的路径拼接bug，导致用户主目录/home/sagemaker-user被重复拼接，目标路径不存在，从而触发FileNotFoundError导致克隆失败。

解决方案 #

1. 手动创建缺失路径，完成克隆

• 打开SageMaker Studio的终端（从Launcher里找到Terminal选项打开）
• 执行命令创建重复路径中的层级：

  mkdir -p /home/sagemaker-user/home/sagemaker-user/cat-ml-test-1-p-mtd5ofsbdgva
  

• 回到克隆操作界面，重新尝试克隆仓库，此时路径已存在，克隆会成功执行。

2. 迁移文件并清理冗余路径（可选，避免后续目录混乱）

克隆成功后，建议把文件迁移到正常路径并清理冗余嵌套目录：

• 先创建正确的项目根目录（如果不存在）：

  mkdir -p /home/sagemaker-user/cat-ml-test-1-p-mtd5ofsbdgva
  

• 把冗余路径里的文件移动到正确目录：

  mv /home/sagemaker-user/home/sagemaker-user/cat-ml-test-1-p-mtd5ofsbdgva/* /home/sagemaker-user/cat-ml-test-1-p-mtd5ofsbdgva/
  

• 删除冗余的嵌套目录：

  rm -rf /home/sagemaker-user/home
  

3. 后续创建项目的规避方法

• 创建CodePipeline MLOps项目时，项目名称尽量简短，不要包含特殊字符
• 在关联克隆仓库前，先在SageMaker Studio终端手动创建对应的项目根目录，再执行克隆操作

错误详情 #

Clone failed Traceback (most recent call last): File "/opt/conda/envs/studio/lib/python3.9/site-packages/jupyterlab_git/git.py", line 144, in execute code, output, error = await current_loop.run_in_executor( File "/opt/conda/envs/studio/lib/python3.9/concurrent/futures/thread.py", line 58, in run result = self.fn(*self.args, **self.kwargs) File "/opt/conda/envs/studio/lib/python3.9/site-packages/jupyterlab_git/git.py", line 110, in call_subprocess process = subprocess.Popen( File "/opt/conda/envs/studio/lib/python3.9/subprocess.py", line 951, in __init__ self._execute_child(args, executable, preexec_fn, close_fds, File "/opt/conda/envs/studio/lib/python3.9/subprocess.py", line 1821, in _execute_child raise child_exception_type(errno_num, err_msg, err_filename)  FileNotFoundError: [Errno 2] No such file or directory: '/home/sagemaker-user/home/sagemaker-user/cat-ml-test-1-p-mtd5ofsbdgva/sagemaker-p-mtd5ofsbdgva-modeldeploy'

内容的提问来源于stack exchange，提问作者dilky97