文档

Creating a Container Image Using a Dockerfile (TensorFlow)

Prerequisites

The software packages of the corresponding OS and the Dockerfile and script files required for packaging images are obtained by referring to Table 1.

In the names of the deep learning engine software package and TensorFlow plugin package, {version} indicates the package version, and {arch} indicates the architecture.

Table 1 Required software

Software Package

Description

How to Obtain

Ascend-cann-nnae_{version}_linux-{arch}.run

Deep learning engine software package.

Link

Ascend-cann-tfplugin_{version}_linux-{arch}.run

Framework plugin package.

Link
  • ARM: tensorflow-{version}-cp37-cp37m-linux_aarch64.whl
  • x86: tensorflow_cpu-{version}-cp37-cp37m-manylinux2010_x86_64.whl

.whl package of the TensorFlow framework.
  • For ARM-based TensorFlow 2.6.5, compile the corresponding .whl package (NumPy version ≥ 1.20.0) by referring to Installing TensorFlow.
  • For x86-based TensorFlow 2.6.5, download the CPU version: pip3 install tensorflow-cpu==2.6.5 from the pip source.

Dockerfile

Required for creating an image.

Prepared by users.

ascend_install.info

Driver installation information file.

Copy the /etc/ascend_install.info file from the host.

version.info

Driver version information file.

Copy the /usr/local/Ascend/driver/version.info file from the host.

prebuild.sh

Script used to prepare for the installation of the training operating environment, for example, configuring the agent.

Prepared by users.

install_ascend_pkgs.sh

Script for installing the Ascend software package.

postbuild.sh

Delete the installation packages, scripts, and proxy configurations that do not need to be retained in the container.

To prevent a software package from being maliciously tampered with during transmission or storage, download the corresponding digital signature file for integrity verification when downloading the software package.

After the software package is downloaded, verify its PGP digital signature according to the OpenPGP Signature Verification Guide. If the software package fails the verification, do not use the software package, and contact Huawei technical support.

Before a software package is used in installation or upgrade, its digital signature also needs to be verified according to OpenPGP Signature Verification Guide to ensure that the software package is not tampered with.

For enterprise users, visit https://support.huawei.com/enterprise/en/tool/pgp-verify-TL1000000054.

This section uses Ubuntu as an example.

Procedure

  1. Upload the software packages, deep learning framework, host driver installation information file, and driver version information file to the same directory (for example, /home/test) on the server.
    • Ascend-cann-nnae_{version}_linux-{arch}.run
    • Ascend-cann-tfplugin_{version}_linux-{arch}.run
    • tensorflow-*_{arch}.whl
    • ascend_install.info
    • version.info
  2. Log in to the server as the root user.
  3. Perform the following steps to prepare the prebuild.sh file:
    1. Go to the directory where the software packages are stored and run the following command to create the prebuild.sh file:

      vim prebuild.sh

    2. For details about the content to be written, see prebuild.sh compilation example. After writing the content, run the :wq command to save the content. The following uses Ubuntu as an example.
  4. Perform the following steps to prepare the install_ascend_pkgs.sh file:
    1. Go to the directory where the software packages are stored and run the following command to create the install_ascend_pkgs.sh file:

      vim install_ascend_pkgs.sh

    2. For details about the content to be written, see install_ascend_pkgs.sh compilation example. After writing the content, run the :wq command to save the content. The following uses Ubuntu as an example.
  5. Perform the following steps to prepare the postbuild.sh file:
    1. Go to the directory where the software packages are stored and run the following command to create the postbuild.sh file:

      vim postbuild.sh

    2. For details about the content to be written, see postbuild.sh compilation example. After writing the content, run the :wq command to save the content. The following uses Ubuntu as an example.
  6. Perform the following steps to create the Dockerfile file:
    1. Go to the directory where the software packages are stored and run the following command to create the Dockerfile file:

      vim Dockerfile

    2. For details about the content to be written, see Dockerfile compilation example. After writing the content, run the :wq command to save the content. The following uses Ubuntu as an example.

      To obtain the image ubuntu:18.04, you can also run the docker pull ubuntu:18.04 command to obtain the image from Docker Hub.

  7. Go to the directory where the software packages are stored and run the following command to create a container image. Do not omit the period (.) at the end of the command.

    docker build -t Image name_System architecture:Image tag .

    Example:

    docker build -t test_train_arm64:v1.0 .

    Table 2 describes the commands.

    Table 2 Parameters

    Parameter

    Description

    -t

    Specifies the image name.

    Image name_System architecture:Image tag

    Image name and tag. Change them based on the actual situation.

    If "Successfully built xxx" is displayed, the image has been created.

  8. After the image is created, run the following command to view the image information:

    docker images

    Example:

    REPOSITORY                TAG                 IMAGE ID            CREATED             SIZE
test_train_arm64          v1.0                d82746acd7f0        27 minutes ago      749MB
  9. Run the following command to access the container:

    docker run -it Image name_System architecture:Image tag bash

    Example:

    docker run -it test_train_arm64:v1.0 bash

  10. Run the following command to obtain the freeze_graph.py file:

    find /usr/local/ -name "freeze_graph.py"

    root@032953231d61:/tmp# find /usr/local/ -name "freeze_graph.py"
/usr/local/lib/python3.7/dist-packages/tensorflow_core/python/tools/freeze_graph.py
  11. Run the following command to modify the file in the image:

    vim /usr/local/lib/python3.7/dist-packages/tensorflow_core/python/tools/freeze_graph.py

    Add the following content.
    from npu_bridge.estimator import npu_ops
from npu_bridge.estimator.npu.npu_config import NPURunConfig
from npu_bridge.estimator.npu.npu_estimator import NPUEstimator
from npu_bridge.estimator.npu.npu_optimizer import allreduce
from npu_bridge.estimator.npu.npu_optimizer import NPUDistributedOptimizer
from npu_bridge.hccl import hccl_ops

    Run the :wq command to save the configuration and exit.

  12. Run the exit command to exit the container.
  13. Run the following command to save the current image:

    docker commit containerid Image name_System architecture:Image tag

    Example:

    root@032953231d61:/tmp# exit
exit
root@ubuntu:/data/kfa/train# docker commit 032953231d61 test_train_arm64:v2.0

    In the preceding example, the value of containerid is 032953231d61.

Compilation Examples

Modify the software package version and architecture based on the actual situation.

  1. Compilation example of prebuild.sh
    • Compilation example of prebuild.sh for the Ubuntu ARM OS
      #!/bin/bash
#--------------------------------------------------------------------------------
# Use the bash syntax to compile script code and prepare for the installation, for example, configuring the proxy.
# This script will be run before the formal creation process is started.
#
# Note: After this script is run, it will not be automatically cleared. If it does not need to be retained in the image, clear it from the postbuild.sh script.
#--------------------------------------------------------------------------------
# DNS settings. If the DNS settings are not required, delete them.
tee /etc/resolv.conf <<- EOF
nameserver xxx.xxx.xxx.xxx # IP address of the DNS server. You can enter multiple IP addresses as required.
nameserver xxx.xxx.xxx.xxx
nameserver xxx.xxx.xxx.xxx
EOF
# APT proxy settings
tee /etc/apt/apt.conf.d/80proxy <<- EOF
Acquire::http::Proxy "http://xxx.xxx.xxx.xxx:xxx";  # IP address and port number of the HTTP proxy server.
Acquire::https::Proxy "http://xxx.xxx.xxx.xxx:xxx";  # IP address and port number of the HTTPS proxy server.
EOF
chmod 777 -R /tmp
rm /var/lib/apt/lists/*
# APT source settings (The following uses Ubuntu 18.04 ARM as an example. Set the information as required.)
tee /etc/apt/sources.list <<- EOF
deb http://mirrors.aliyun.com/ubuntu-ports/ bionic main restricted universe multiverse
deb-src http://mirrors.aliyun.com/ubuntu-ports/ bionic main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu-ports/ bionic-security main restricted universe multiverse
deb-src http://mirrors.aliyun.com/ubuntu-ports/ bionic-security main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu-ports/ bionic-updates main restricted universe multiverse
deb-src http://mirrors.aliyun.com/ubuntu-ports/ bionic-updates main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu-ports/ bionic-proposed main restricted universe multiverse
deb-src http://mirrors.aliyun.com/ubuntu-ports/ bionic-proposed main restricted universe multiverse
deb http://mirrors.aliyun.com/ubuntu-ports/ bionic-backports main restricted universe multiverse
deb-src http://mirrors.aliyun.com/ubuntu-ports/ bionic-backports main restricted universe multiverse
EOF
    • Compilation example of prebuild.sh for the Ubuntu x86 OS
      #!/bin/bash
#--------------------------------------------------------------------------------

# Use the bash syntax to compile script code and prepare for the installation, for example, configuring the proxy.
# This script will be run before the formal creation process is started.
#
# Note: After this script is run, it will not be automatically cleared. If it does not need to be retained in the image, clear it from the postbuild.sh script.
#--------------------------------------------------------------------------------
# APT proxy settings
tee /etc/apt/apt.conf.d/80proxy <<- EOF
Acquire::http::Proxy "http://xxx.xxx.xxx.xxx:xxx";    #IP address and port number of the HTTP proxy server.
Acquire::https::Proxy "http://xxx.xxx.xxx.xxx:xxx";  #IP address and port number of the HTTPS proxy server.
EOF

#APT source settings (The following uses Ubuntu 18.04 x86 as an example. Set the information based on the site requirements.)
tee /etc/apt/sources.list <<- EOF
deb http://mirrors.ustc.edu.cn/ubuntu/ bionic main multiverse restricted universe
deb http://mirrors.ustc.edu.cn/ubuntu/ bionic-backports main multiverse restricted universe
deb http://mirrors.ustc.edu.cn/ubuntu/ bionic-proposed main multiverse restricted universe
deb http://mirrors.ustc.edu.cn/ubuntu/ bionic-security main multiverse restricted universe
deb http://mirrors.ustc.edu.cn/ubuntu/ bionic-updates main multiverse restricted universe
deb-src http://mirrors.ustc.edu.cn/ubuntu/ bionic main multiverse restricted universe
deb-src http://mirrors.ustc.edu.cn/ubuntu/ bionic-backports main multiverse restricted universe
deb-src http://mirrors.ustc.edu.cn/ubuntu/ bionic-proposed main multiverse restricted universe
deb-src http://mirrors.ustc.edu.cn/ubuntu/ bionic-security main multiverse restricted universe
deb-src http://mirrors.ustc.edu.cn/ubuntu/ bionic-updates main multiverse restricted universe
EOF
  2. Compilation example of install_ascend_pkgs.sh
    #!/bin/bash
#--------------------------------------------------------------------------------
# Use the bash syntax to compile script code and install the Ascend software package.
#
# Note: After this script is run, it will not be automatically cleared. If it does not need to be retained in the image, clear it from the postbuild.sh script.
#--------------------------------------------------------------------------------
# Copy the /etc/ascend_install.info file on the host to the current directory before creating the container image.
cp ascend_install.info /etc/
# Copy the /usr/local/Ascend/driver/version.info file on the host to the current directory before creating the container image.
mkdir -p /usr/local/Ascend/driver/
cp version.info /usr/local/Ascend/driver/
 
# Ascend-cann-nnae_{version}_linux-{arch}.run
chmod +x Ascend-cann-nnae_{version}_linux-{arch}.run
./Ascend-cann-nnae_{version}_linux-{arch}.run --install-path=/usr/local/Ascend/ --install --quiet
 
# Ascend-cann-tfplugin_{version}_linux-{arch}.run
chmod +x Ascend-cann-tfplugin_{version}_linux-{arch}.run
./Ascend-cann-tfplugin_{version}_linux-{arch}.run  --install --quiet
 
# After the NNAE package is installed, clear the following files. During container startup, the NNAE package is mounted by the Ascend Docker.
rm -f version.info
rm -rf /usr/local/Ascend/driver/

    If the following information is displayed during image creation, delete parameter --install-path following Ascend-cann-xxx.run (except when the first Ascend-cann-xxx.run package is installed):

    • Information displayed:
      [TFPlugin] [20210316-02:39:37] [ERROR] /etc/Ascend/ascend_cann_install.info exists ! 'install-path' parameter are not supported.
    • Possible causes:

      After the first CANN software package is installed, the installation path is recorded in the /etc/Ascend/ascend_cann_install.info file. If this file exists, it will be automatically installed in the path recorded in this file when other CANN software packages are installed. In this case, the --install-path parameter is not supported.

  3. Compilation example of postbuild.sh (Ubuntu)
    #!/bin/bash
#--------------------------------------------------------------------------------
# Use the bash syntax to compile the script code and delete the installation packages, scripts, and proxy configurations that do not need to be retained in the container.
# This script will be run after the formal creation process ends.
#
# Note: After this script is run, it is automatically cleared and will not be left in the image. The scripts and Working Dir are stored in /root.
#--------------------------------------------------------------------------------

rm -f ascend_install.info
rm -f prebuild.sh
rm -f install_ascend_pkgs.sh
rm -f Dockerfile
rm -f Ascend-cann-nnae_{version}_linux-{arch}.run
rm -f Ascend-cann-tfplugin_{version}_linux-{arch}.run
# ARM environment
rm -f tensorflow-1.15.0-cp37-cp37m-linux_{arch}.whl
# If the offline package is used for installation in the x86 environment, comment out the previous line and delete the comment tag (#) from the next line.
# rm -f tensorflow_cpu-1.15.0-cp37-cp37m-manylinux2010_x86_64.whl
rm -f /etc/apt/apt.conf.d/80proxy
 
# Delete if not required
tee /etc/resolv.conf <<- EOF
# This file is managed by man:systemd-resolved(8). Do not edit.
#
# This is a dynamic resolv.conf file for connecting local clients to the
# internal DNS stub resolver of systemd-resolved. This file lists all
# configured search domains.
#
# Run systemd-resolve --status to see details about the uplink DNS servers
# currently in use.
#
# Third party programs must not access this file directly, but only through the
# symlink at /etc/resolv.conf. To manage man:resolv.conf(5) in a different way,
# replace this symlink by a static file or a different symlink.
#
# See man:systemd-resolved.service(8) for details about the supported modes of
# operation for /etc/resolv.conf.
 
options edns0
 
nameserver xxx.xxx.xxx.xxx
nameserver xxx.xxx.xxx.xxx
EOF
  4. Dockerfile compilation sample
    • Dockerfile example for Ubuntu ARM
      FROM ubuntu:18.04

ARG TF_PKG=tensorflow-1.15.0-cp37-cp37m-linux_aarch64.whl
ARG HOST_ASCEND_BASE=/usr/local/Ascend
ARG NNAE_PATH=/usr/local/Ascend/nnae/latest
ARG TF_PLUGIN_PATH=/usr/local/Ascend/tfplugin/latest
ARG INSTALL_ASCEND_PKGS_SH=install_ascend_pkgs.sh
ARG PREBUILD_SH=prebuild.sh
ARG POSTBUILD_SH=postbuild.sh
WORKDIR /tmp
COPY . ./

# Trigger prebuild.sh.
RUN bash -c "test -f $PREBUILD_SH && bash $PREBUILD_SH || true"

ENV http_proxy http://xxx.xxx.xxx.xxx:xxx
ENV https_proxy http://xxx.xxx.xxx.xxx:xxx

# System package
RUN apt update && \
    apt install --no-install-recommends \
        python3.7 python3.7-dev \
        curl g++ pkg-config unzip \
        libblas3 liblapack3 liblapack-dev \
        libblas-dev gfortran libhdf5-dev \
        libffi-dev libicu60 libxml2 -y

# Create a Python soft link.
RUN ln -s /usr/bin/python3.7 /usr/bin/python
# Configure the Python pip source.
RUN mkdir -p ~/.pip \
&& echo '[global] \n\
index-url=https://pypi.doubanio.com/simple/\n\
trusted-host=pypi.doubanio.com' >> ~/.pip/pip.conf

# pip3.7
RUN curl -k https://bootstrap.pypa.io/get-pip.py -o get-pip.py && \
    cd /tmp && \
    apt-get download python3-distutils && \
    dpkg-deb -x python3-distutils_*.deb / && \
    rm python3-distutils_*.deb && \
    cd - && \
    python3.7 get-pip.py && \
    rm get-pip.py

# Create the HwHiAiUser user and owner. The values of UID and GID must be the same as those on the physical machine to avoid generating ownerless files. In the example, the user and the corresponding group are automatically created. The values of UID and GID are both 1000.
RUN useradd -d /home/HwHiAiUser -u 1000 -m -s /bin/bash HwHiAiUser

# Python package
RUN pip3.7 install numpy && \
    pip3.7 install decorator && \
    pip3.7 install sympy==1.4 && \
    pip3.7 install cffi && \
    pip3.7 install pyyaml && \
    pip3.7 install pathlib2 && \
    pip3.7 install grpcio && \
    pip3.7 install grpcio-tools && \
    pip3.7 install protobuf && \
    pip3.7 install scipy && \
    pip3.7 install requests

# Ascend package
RUN umask 0022 && bash $INSTALL_ASCEND_PKGS_SH

RUN umask 0022 && pip3.7 install $TF_PKG

# Create /lib64/ld-linux-aarch64.so.1.
RUN umask 0022 && \
    if [ ! -d "/lib64" ]; \
    then \
        mkdir /lib64 && ln -sf /lib/ld-linux-aarch64.so.1 /lib64/ld-linux-aarch64.so.1; \
    fi

ENV http_proxy ""
ENV https_proxy ""

# Trigger postbuild.sh.
RUN bash -c "test -f $POSTBUILD_SH && bash $POSTBUILD_SH || true" && \
    rm $POSTBUILD_SH
    • Dockerfile example for Ubuntu x86
      FROM ubuntu:18.04
# The following lines are used for online download and installation during image compilation. It is mutually exclusive with the .whl configuration.
ARG TF_PKG=tensorflow-cpu==1.15.0
# Use the offline x86 TensorFlow package, comment out the upper line, and delete the comment tag (#) from the lower line.
#ARG TF_PKG=tensorflow_cpu-1.15.0-cp37-cp37m-manylinux2010_x86_64.whl
ARG HOST_ASCEND_BASE=/usr/local/Ascend
ARG NNAE_PATH=/usr/local/Ascend/nnae/latest
ARG TF_PLUGIN_PATH=/usr/local/Ascend/tfplugin/latest
ARG INSTALL_ASCEND_PKGS_SH=install_ascend_pkgs.sh
ARG PREBUILD_SH=prebuild.sh
ARG POSTBUILD_SH=postbuild.sh
WORKDIR /tmp
COPY . ./

# Trigger prebuild.sh.
RUN bash -c "test -f $PREBUILD_SH && bash $PREBUILD_SH || true"

ENV http_proxy http://xxx.xxx.xxx.xxx:xxx
ENV https_proxy http://xxx.xxx.xxx.xxx:xxx

# System package
RUN apt update && \
    apt install --no-install-recommends \
        python3.7 python3.7-dev \
        curl g++ pkg-config unzip \
        libblas3 liblapack3 liblapack-dev \
        libblas-dev gfortran libhdf5-dev \
        libffi-dev libicu60 libxml2 -y

# Create a Python soft link.
RUN ln -s /usr/bin/python3.7 /usr/bin/python

# Configure the Python pip source.
RUN mkdir -p ~/.pip \
&& echo '[global] \n\
index-url=https://pypi.doubanio.com/simple/\n\
trusted-host=pypi.doubanio.com' >> ~/.pip/pip.conf

# pip3.7
RUN curl -k https://bootstrap.pypa.io/get-pip.py -o get-pip.py && \
    cd /tmp && \
    apt-get download python3-distutils && \
    dpkg-deb -x python3-distutils_*.deb / && \
    rm python3-distutils_*.deb && \
    cd - && \
    python3.7 get-pip.py && \
    rm get-pip.py

# Create the HwHiAiUser user and owner. The values of UID and GID must be the same as those on the physical machine to avoid generating ownerless files. In the example, the user and the corresponding group are automatically created. The values of UID and GID are both 1000.
RUN useradd -d /home/HwHiAiUser -u 1000 -m -s /bin/bash HwHiAiUser

# Python package
RUN pip3.7 install numpy && \
    pip3.7 install decorator && \
    pip3.7 install sympy==1.4 && \
    pip3.7 install cffi==1.12.3 && \
    pip3.7 install pyyaml && \
    pip3.7 install pathlib2 && \
    pip3.7 install grpcio && \
    pip3.7 install grpcio-tools && \
    pip3.7 install protobuf && \
    pip3.7 install scipy && \
    pip3.7 install requests

# Ascend package
RUN umask 0022 && bash $INSTALL_ASCEND_PKGS_SH

RUN pip3.7 install $TF_PKG

ENV http_proxy ""
ENV https_proxy ""

# Trigger postbuild.sh.
RUN bash -c "test -f $POSTBUILD_SH && bash $POSTBUILD_SH || true" && \
    rm $POSTBUILD_SH
Parent topic: Common Operations