linux - 构建docker:找不到opt/conda/bin/conda

我是docker的新手，我需要在docker环境中运行代码。
我在构建Dockerfile时出错:
我正在通过hyper-V运行Ubuntu 20.04，并且在构建Dockerfile时，收到以下消息:

Step 4/20 : RUN curl -o ~/miniconda.sh -O  https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh  &&      chmod +x ~/miniconda.sh &&      ~/miniconda.sh -b -p /opt/conda &&      rm ~/miniconda.sh &&      /opt/conda/bin/conda install numpy pyyaml scipy ipython mkl &&      /opt/conda/bin/conda install -c soumith magma-cuda90 &&      /opt/conda/bin/conda clean -ya <br />
 ---> Running in 9758f4fe60a4 <br />
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current

                                 Dload  Upload   Total   Spent    Left  Speed
  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
/bin/sh: 1: /opt/conda/bin/conda: not found
The command '/bin/sh -c curl -o ~/miniconda.sh -O  https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh  &&      chmod +x ~/miniconda.sh &&      ~/miniconda.sh -b -p /opt/conda &&      rm ~/miniconda.sh &&      /opt/conda/bin/conda install numpy pyyaml scipy ipython mkl &&      /opt/conda/bin/conda install -c soumith magma-cuda90 &&      /opt/conda/bin/conda clean -ya' returned a non-zero code: 127

Dockerfile的:

# PyTorch Install
FROM nvidia/cuda:9.0-cudnn7-devel-ubuntu16.04

RUN apt-get update && apt-get install -y --no-install-recommends \
         build-essential \
         cmake \
         git \
         curl \
         vim \
     emacs \
         parallel \
         ca-certificates \
         libjpeg-dev \
     hdf5-tools \
         libpng-dev &&\
     rm -rf /var/lib/apt/lists/*


RUN mkdir ~/.parallel && touch ~/.parallel/will-cite

RUN sudo curl -o ~/miniconda.sh -O  https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh  && \
     chmod +x ~/miniconda.sh && \
     ~/miniconda.sh -b -p /opt/conda && \
     rm ~/miniconda.sh && \
     /opt/conda/bin/conda install numpy pyyaml scipy ipython mkl && \
     /opt/conda/bin/conda install -c soumith magma-cuda90 && \
     /opt/conda/bin/conda clean -ya
ENV PATH /opt/conda/bin:$PATH
# This must be done before pip so that requirements.txt is available
WORKDIR /opt/pytorch
COPY . .
RUN conda install pytorch torchvision -c pytorch

#
# Now install the julia dependencies.
#
WORKDIR /opt/julia
RUN pip install pandas matplotlib utils argh biopython
RUN conda install networkx joblib

RUN apt-get update && apt-get install -y curl
RUN mkdir /julia
RUN curl -L https://julialang-s3.julialang.org/bin/linux/x64/0.6/julia-0.6.2-linux-x86_64.tar.gz | tar -C /julia --strip-components=1  -xzf -
ENV PATH "/julia/bin:$PATH"
RUN julia -e "Pkg.init()"
COPY setup.jl /julia/setup.jl
RUN julia /julia/setup.jl

WORKDIR /root/hyperbolics
ENV PYTHONPATH /root/hyperbolics

当我在命令提示符下直接在RUN下尝试命令时，它运行良好。
但是，当我构建Dockerfile，julia，pip等时，每个命令都是'not found'。(当我注释掉conda ...部分时)
我怎么解决这个问题？

最佳答案

/bin/sh: 1: /opt/conda/bin/conda: not found错误是由于未正确安装conda引起的。这是因为您下载的miniconda.sh文件(带有curl -o ~/miniconda.sh -O https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh)是一个空文件。
发生这种情况是因为curl by default doesn't follow redirects和来自上面url的响应发送了重定向，而不是直接发送miniconda.sh文件。您可以通过检查网址的标题来验证这一点(例如curl -i https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh显示HTTP 301重定向状态代码)。
您可以通过在curl命令中提供-L标志来告诉curl遵循重定向，从而解决此问题。 :
(我还必须删除sudo)

RUN curl -L -o ~/miniconda.sh -O  https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh  && \
     chmod +x ~/miniconda.sh && \
     ~/miniconda.sh -b -p /opt/conda && \
     rm ~/miniconda.sh && \
     /opt/conda/bin/conda install numpy pyyaml scipy ipython mkl && \
     /opt/conda/bin/conda install -c soumith magma-cuda90 && \
     /opt/conda/bin/conda clean -ya

这是一些调试信息/研究:
您可以通过查看以上代码段中curl的输出来验证空白miniconda.sh是问题所在。看起来有点神秘，但您实际上什么都没有下载

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current

                                 Dload  Upload   Total   Spent    Left  Speed
  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0

当我尝试构建添加了-L标志的docker镜像时，输出如下所示:

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
  0     0    0     0    0     0      0      0 --:--:--  0:00:03 --:--:--     0
100 88.7M  100 88.7M    0     0  11.9M      0  0:00:07  0:00:07 --:--:-- 34.7M

您还可以通过删除下载/运行部分之前的所有内容来手动检查miniconda.sh的内容
(例如，将所有内容都保留到这一部分)

RUN sudo curl -o ~/miniconda.sh -O  https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh  && \
     chmod +x ~/miniconda.sh && \
     ~/miniconda.sh -b -p /opt/conda

然后运行sudo docker run -it containername /bin/bash并检查miniconda.sh的内容(cat ~/miniconda.sh，它显示文件为空)。

关于linux - 构建docker:找不到opt/conda/bin/conda，我们在Stack Overflow上找到一个类似的问题：https://stackoverflow.com/questions/63359198/