none
Docker image Build failed. status code 400 RRS feed

  • Question

  • last month, I was able to build the docker image using the jupyter notebook.

    But this today, I received this message despite putting pip in the pip_packages setting.

    "Downloading and Extracting Packages

    ca-certificates-2019 | 145 KB    |            |   0%
    ca-certificates-2019 | 145 KB    | ########## | 100%
    certifi-2019.11.28   | 149 KB    |            |   0%
    certifi-2019.11.28   | 149 KB    | ########## | 100%
    pip-20.0.1           | 1.9 MB    |            |   0%
    pip-20.0.1           | 1.9 MB    | #######8   |  79%
    pip-20.0.1           | 1.9 MB    | #########6 |  97%
    pip-20.0.1           | 1.9 MB    | ########## | 100%
    openssl-1.1.1d       | 2.1 MB    |            |   0%
    openssl-1.1.1d       | 2.1 MB    | ########   |  80%
    openssl-1.1.1d       | 2.1 MB    | ########## | 100%

    Pip subprocess error:

    ERROR: Could not find a version that satisfies the requirement dotnetcore2>=2.1.9 (from azureml-dataprep[fuse]<1.2.0a,>=1.1.20a->azureml-sdk==1.0.69.*->-r /var/azureml-app/condaenv.d2xx28n3.requirements.txt (line 1)) (from versions: none)
    ERROR: No matching distribution found for dotnetcore2>=2.1.9 (from azureml-dataprep[fuse]<1.2.0a,>=1.1.20a->azureml-sdk==1.0.69.*->-r /var/azureml-app/condaenv.d2xx28n3.requirements.txt (line 1))"

    There appears to be this issue with pip.

    How can I change the pip to a lower version?

    This is the section where I specified the lower version but to no avail.

    from azureml.core.conda_dependencies import CondaDependencies
    myenv = CondaDependencies.create(pip_packages=['azureml-sdk','azure-storage','numpy','pandas','azure-cosmos','pip==19.0'])
    with open("myenv.yml","w") as f:
        f.write(myenv.serialize_to_string())
    Thursday, January 23, 2020 12:57 AM

Answers

  • Hello,

    Yes, This is a known issue and our team is working on it. Please refer the following issue for more details.

    We would recommend to use the workaround steps from the GIT issue to mitigate the current error.

    -----------------------------------------------------------------------------------------------------------
    If you found this post helpful, please give it a "Helpful" vote. 
    Please remember to mark the replies as answers if they help. 

    Thursday, January 30, 2020 3:02 PM
    Moderator

All replies

  • Hello,

    You can try to install the correct version pip in your notebook environment before running any of the steps above. Using this in one of the cell should correct it but if there is a dependency for a higher version in any of the other packages it might fail again.

    !pip install pip==19.0

    Hope this helps!!

    -Rohit

    Thursday, January 23, 2020 5:50 AM
    Moderator
  • Thanks for the reply.

    I used "pip==19.3.1" in the cell. It seemed to download that version but if you notice, it was also downloading pip-20.0.1. So the 19.3.1 probably got overwritten by 20.0.1 or so I think.

    Downloading and Extracting Packages
    ca-certificates-2019 | 145 KB    |            |   0%
    ca-certificates-2019 | 145 KB    | ########## | 100%
    certifi-2019.11.28   | 149 KB    |            |   0%
    certifi-2019.11.28   | 149 KB    | ########## | 100%
    pip-20.0.1           | 1.9 MB    |            |   0%
    pip-20.0.1           | 1.9 MB    | #######8   |  79%
    pip-20.0.1           | 1.9 MB    | ########## | 100%
    openssl-1.1.1d       | 2.1 MB    |            |   0%
    openssl-1.1.1d       | 2.1 MB    | #######9   |  80%
    openssl-1.1.1d       | 2.1 MB    | ########## | 100%
    Preparing transaction: ...working... done
    Verifying transaction: ...working... done
    Executing transaction: ...working... done
    
    CondaEnvException: Pip failed
    Ran pip subprocess with arguments:
    ['/opt/miniconda/bin/python', '-m', 'pip', 'install', '-U', '-r', '/var/azureml-app/condaenv.6w_lh1um.requirements.txt']
    Pip subprocess output:
    Collecting azureml-sdk==1.0.69.*
      Downloading azureml_sdk-1.0.69-py3-none-any.whl (4.6 kB)
    Collecting azure-storage
      Downloading azure_storage-0.36.0-py2.py3-none-any.whl (190 kB)
    Collecting numpy
      Downloading numpy-1.18.1-cp36-cp36m-manylinux1_x86_64.whl (20.1 MB)
    Collecting pandas
      Downloading pandas-0.25.3-cp36-cp36m-manylinux1_x86_64.whl (10.4 MB)
    Collecting azure-cosmos
      Downloading azure_cosmos-3.1.2-py3-none-any.whl (140 kB)
    Collecting pip==19.3.1
      Downloading pip-19.3.1-py2.py3-none-any.whl (1.4 MB)

    ...

    Pip subprocess error:

    ERROR: Could not find a version that satisfies the requirement dotnetcore2>=2.1.9 (from azureml-dataprep[fuse]<1.2.0a,>=1.1.20a->azureml-sdk==1.0.69.*->-r /var/azureml-app/condaenv.6w_lh1um.requirements.txt (line 1)) (from versions: none)
    ERROR: No matching distribution found for dotnetcore2>=2.1.9 (from azureml-dataprep[fuse]<1.2.0a,>=1.1.20a->azureml-sdk==1.0.69.*->-r /var/azureml-app/condaenv.6w_lh1um.requirements.txt (line 1))
    The command '/bin/sh -c CONDA_ROOT_DIR=$(conda info --root) && if [ -n "$AZUREML_CONDA_ENVIRONMENT_PATH" ]; then conda env update -p "$AZUREML_CONDA_ENVIRONMENT_PATH" -f '/var/azureml-app/myenv.yml'; else conda env update -n base -f '/var/azureml-app/myenv.yml'; fi && conda clean -aqy && rm -rf /root/.cache/pip && rm -rf "$CONDA_ROOT_DIR/pkgs" && find "$CONDA_ROOT_DIR" -type d -name __pycache__ -exec rm -rf {} +' returned a non-zero code: 1
    2020/01/23 22:08:13 Container failed during run: acb_step_0. No retries remaining.
    failed to run step ID: acb_step_0: exit status 1
    Run ID: da3s failed after 1m4s. Error: failed during run, err: exit status 1

    • Edited by kfandst Thursday, January 23, 2020 10:17 PM
    Thursday, January 23, 2020 9:43 PM
  • Hello,

    Yes, This is a known issue and our team is working on it. Please refer the following issue for more details.

    We would recommend to use the workaround steps from the GIT issue to mitigate the current error.

    -----------------------------------------------------------------------------------------------------------
    If you found this post helpful, please give it a "Helpful" vote. 
    Please remember to mark the replies as answers if they help. 

    Thursday, January 30, 2020 3:02 PM
    Moderator