92 篇文章带有标签 “docker”

Docker SDK for Python Examples

安装 Docker SDK for Python

pip install docker

例子

将本地文件或目录添加到容器，生成新的镜像。

  • put_archive
  • commit
import docker
import tarfile
import tempfile
import os

def simple_tar(path):
    f = tempfile.NamedTemporaryFile()
    t = tarfile.open(mode='w', fileobj=f)
    abs_path = os.path.abspath(path)
    t.add(abs_path, arcname=os.path.basename(path))
    t.close()
    f.seek(0)
    return f

client = docker.from_env()
// ...

参考资料

dockerpythonsdkcontainer-managementtempfilescripting

如何使用 Docker 打包已注册的模型

模型打包成镜像

手工打包

project_dir=platen-switch
darknet_model_name=darknet-model-platen-switch

cd $project_dir
docker run -d --name $darknet_model_name alpine
docker cp model/ $darknet_model_name:/
docker commit -a 'wang-junjian@qq.com' -m 'darknet model [platen-switch recognition]' \
  $darknet_model_name gouchicao/$darknet_model_name:latest
docker rm -v $darknet_model_name

docker push gouchicao/$darknet_model_name:latest

Python 脚本 import docker import tarfile import tempfile import os def simple_tar(path): f = tempfile.NamedTemporaryFile() t = tarfile.

dockerkubernetesmodel-deploymentinference-servingdarknetinit-containerspythonsdkcontainer-management

minikube

安装

安装 kubectl

curl -LO "https://storage.googleapis.com/kubernetes-release/release/$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)/bin/linux/amd64/kubectl"
chmod +x ./kubectl
mv ./kubectl /usr/local/bin/kubectl

安装 minikube

curl -LO https://storage.googleapis.com/minikube/releases/latest/minikube-linux-amd64
install minikube-linux-amd64 /usr/local/bin/minikube

启动 Kubernetes 集群 # minikube start --force 😄 Centos 8.1.1911 (amd64) 上的 minikube v1.22.

kuberneteskubectldockerminikubecluster-setuplocal-developmentaddonscentos

ConfigMap和Secret：配置应用程序

ConfigMap

向容器传递命令行参数

在 Docker 中定义命令与参数

下面是 Dockerfile 中的指令 ENTRYPOINT 和 CMD。

指令 解释
ENTRYPOINT 容器启动时调用的命令
CMD 传递给 ENTRYPOINT 指定命令的参数

构建带参数的程序（date:args）

编写脚本 date.sh

#!/bin/sh
INTERVAL=$1
while :
do
  echo $(date)
  sleep $INTERVAL
done

编写 Dockerfile

FROM busybox
ADD date.sh /date.sh
RUN chmod +x /date.sh
ENTRYPOINT ["/date.sh"]
CMD ["1"]

参数默认值为 1，在运行容器可以设置参数覆盖默认值。

构建镜像

docker build -t wangjunjian/date:args .

在 Docker 中覆盖命令和参数

使用 Docker 运行镜像来设置命令和参数

docker run [--entrypoint=] <image> [arg1, arg2, arg3]

启动镜像 $ docker run wangjunjian/date:args Tue Aug 3 13:08:04 UTC 2021 Tue Aug 3

kubernetesconfigmapsecretdockerdockerfilenginxhttpsenvironment-variablesvolumesport-forward

Kubernetes中的卷：将磁盘挂载到容器

通过卷在容器之间共享数据

Build [Date HTML Generator] Image

编写 HTML 生成器（date-html-generator.sh）

#!/bin/sh
mkdir /var/htdocs
while :
do
  echo $(date +'%Y-%m-%d %H:%M:%S') Writing to /var/htdocs/index.html
  echo $(date +'%Y-%m-%d %H:%M:%S') > /var/htdocs/index.html
  sleep 1
done
  • %Y : 完整年份 (0000-9999)
  • %m : 月份 (01-12)
  • %d : 日 (01-31)
  • %H : 小时(00-23)
  • %M : 分钟(00-59)
  • %S : 秒(00-60)

编写 Dockerfile

FROM busybox
ADD date-html-generator.sh /bin/date-html-generator.sh
RUN chmod +x /bin/date-html-generator.sh
ENTRYPOINT /bin/date-html-generator.sh

生成 date-html-generator 镜像 docker build -t wangjunjian/date-htm

kubernetesvolumesstorageemptydirhostpathnfspersistent-storagedockerport-forward

在Kubernetes上运行第一个应用

基于Node构建应用

app.js

const http = require('http');
const os = require('os');

console.log("Kubia server starting...");

var handler = function(request, response) {
  console.log("Received request from " + request.connection.remoteAddress);
  response.writeHead(200);
  response.end("You've hit " + os.hostname() + "\n");
};

var www = http.createServer(handler);
www.listen(8080);

Dockerfile

FROM node:16-slim
ADD app.js /app.js
ENTRYPOINT ["node", "app.js"]

构建

docker build -t wangjunjian/kubia .

运行 docker run -d -p 8080:8080 wangjun

kubernetesdockerdockerfilenodejsservicedeploymentnodeportbeginner

Kubernetes中的GPU共享

构建应用

Scheduler Extender

git clone https://github.com/AliyunContainerService/gpushare-scheduler-extender.git && cd gpushare-scheduler-extender
docker build -t gouchicao/gpushare-scheduler-extender .

Device Plugin

git clone https://github.com/AliyunContainerService/gpushare-device-plugin.git && cd gpushare-device-plugin
docker build -t gouchicao/gpushare-device-plugin .

Kubectl Extension

wget https://github.com/AliyunContainerService/gpushare-device-plugin/releases/download/v0.3.0/kubectl-inspect-gpushare

安装 在控制平面中部署 GPU 共享调度程序扩展器 cd /etc/kubernetes sudo wget https://raw.

kubernetesgpunvidiadockerdevice-pluginschedulinggpu-sharingaliyun

AI 模型打包发布

工程目录

model-package-release/
├── Dockerfile
├── Dockerfile.ubuntu
├── main.sh
└── .ssh
    ├── id_rsa
    ├── id_rsa.pub
    └── known_hosts

main.sh #!/usr/bin/sh # 模型和配置文件 FILE_CONFIG=config.yaml FILE_MODEL=model.onnx # 用于构建模型压缩包的目录结构 DIR_MODELS=models if [ -z $1 ] then echo "Usage: " echo "Environment variable：" echo " MODEL_SERVER_USERNAME Default: username" echo " MODEL_SERVER_IP Default: ip" echo "" echo "docker run --rm -v /home/ai/models/sign.yaml:/app/config.yaml \" echo " -v /home/ai/models/sign.onnx:/app/model.

aimachine-learningdockerdockerfiledocker-composeshellonnxmodel-deploymentpackaging

容器内幕

容器的本质是一种特殊的进程

  • Linux Namespace - 空间隔离
  • Linux Cgroups - 资源限制
  • chroot - 切换进程的根目录

Linux Namespace

Linux Cgroups - Linux Control Group

作用：限制一个进程组能够使用的资源上限，包括 CPU、内存、磁盘、网络带宽等等。

查看Cgroups限制的资源各类

在 Linux 中，Cgroups 给用户暴露出来的操作接口是文件系统，即它以文件和目录的方式组织在操作系统的 /sys/fs/cgroup 路径下。可以用 mount 指令把它们展示出来。

containerslinux-kernelcgroupsnamespaceschrootdockervirtualizationlinux

Install NVIDIA device plugin for Kubernetes

配置每个NVIDIA GPU节点上的Docker

  1. 增加"default-runtime": "nvidia"
$ sudo vim /etc/docker/daemon.json
{
    "registry-mirrors": ["https://75oltije.mirror.aliyuncs.com"],
    "default-runtime": "nvidia",
    "runtimes": {
        "nvidia": {
            "path": "nvidia-container-runtime",
            "runtimeArgs": []
        }
    }
}
  1. 重启服务
sudo systemctl restart docker

设置每个节点的污点

GPU 节点

kubectl taint node gpu1 nvidia.com/gpu:NoSchedule
kubectl taint node gpu2 nvidia.com/gpu:NoSchedule

CPU 节点 kubectl taint node ln2 node-type=production:NoSchedule kubectl ta

kubernetesnvidiagpudockerhelmubuntukubectlnvidia-docker2device-plugin

基于PyPIServer创建私有Python软件包存储库

服务端

拉取PyPIServer镜像

docker pull pypiserver/pypiserver:latest

部署PyPIServer

  • 只用于客户端下载（用作缓存加速）
docker run -d --restart=always --name pypiserver -p 8080:8080 \
    -v /data/pypi-packages/:/data/packages \
    pypiserver/pypiserver:latest
  • 客户端不仅可以下载还可以上传（当我们自己开发了Python的软件时）
#创建用户名和密码
sudo apt install apache2-utils -y
sudo mkdir /data/pypi-packages
sudo htpasswd -sc /data/pypi-packages/htpasswd.txt wjj
#当您需要再创建用户名时就不需要加参数 -c
sudo htpasswd -s /data/pypi-packages/htpasswd.txt test
#容器部署
docker run -d --restart=always --name pypiserver -p 8080:8080 \
    -v /data/pypi-packages/:/data/packages \
    pypiserver/pypiserver:latest -P /data/packages/htpasswd.txt
pythonpippypipypiserverdockerhtpasswdrepositorypackagingprivate-registrydevops

Docker实践

安装与卸载

安装

  • 快速安装
curl -fsSL https://get.docker.com | sh -

卸载

  • apt
apt-get remove --auto-remove docker
  • yum
yum remove docker docker-common docker-selinux docker-engine

指定显卡(NVIDIA_VISIBLE_DEVICES)

#指定单张GPU卡
docker run --runtime=nvidia -d -e NVIDIA_VISIBLE_DEVICES=0 gouchicao/yolov5:train
#指定多张GPU卡，多个逗号隔开。
docker run --runtime=nvidia -d -e NVIDIA_VISIBLE_DEVICES=0,1 gouchicao/yolov5:train
#NVIDIA_VISIBLE_DEVICES默认值是all
docker run --runtime=nvidia -d gouchicao/yolov5:train

重启策略 方法 启动容器时通过参数指定 #如果容器停止总是重新启动。如果手动停止，则仅在Docker守护程序重启或手动重启容器本身时才重启。

dockergpunvidiacontainer-managementcontainersnvidia-docker2devopsdeploymentlinuxtips

Dockerfile OpenCV4 Ubuntu20.04

Dockerfile

FROM ubuntu:20.04
LABEL maintainer="wang-junjian@qq.com"

#auto install tzdata(opencv depend)
ENV DEBIAN_FRONTEND=noninteractive

RUN sed -i s@/archive.ubuntu.com/@/mirrors.aliyun.com/@g /etc/apt/sources.list
RUN apt-get update && apt-get install -y \
    python3 python3-pip \
    libglib2.0-dev libgl1-mesa-glx \
    && rm -rf /var/lib/apt/lists/*

RUN ln -s /usr/bin/python3 /usr/bin/python && \
    ln -s /usr/bin/pip3 /usr/bin/pip
RUN pip install --no-cache-dir -i https://mirrors.aliyun.com/pypi/simple/ \
    opencv-python opencv-contrib-python

#set your localtime
RUN ln -fs /usr/share/zoneinfo/Asia/Shanghai /etc/localtime

WORKDIR /

CMD bash
dockerdockerfileopencvubuntucomputer-visionimage-processingpythoncontainerstips

Building ONNX Runtime

NVIDIA CUDA

单步构建

  • 下载onnxruntime源代码
git clone --recursive https://github.com/microsoft/onnxruntime.git
  • 拉取容器（编译环境）
docker pull nvidia/cuda:11.1-cudnn8-devel-ubuntu20.04
  • 运行容器
docker run -it --name build-onnxruntime-gpu --runtime nvidia \
    -v $(pwd)/onnxruntime:/onnxruntime -w /onnxruntime \
    nvidia/cuda:11.1-cudnn8-devel-ubuntu20.04
  • 更新apt镜像源
sed -i 's/archive.ubuntu.com/mirrors.aliyun.com/g' /etc/apt/sources.list
apt-get update
  • 安装依赖包
apt-get install language-pack-en git cmake python3 python3-pip -y
  • 修改语言环境
locale-gen en_US.UTF-8
update-locale LANG=en_US.UTF-8
  • 更新pip镜像源
pip3 config set global.index-url https://mirrors.aliyun.com/pypi/simple/
linuxubuntugpucudadockeronnxonnxruntimenvidiadeep-learningbuild-from-source

Dockerfile ONNXRuntime GPU 

FROM nvidia/cuda:11.1-cudnn8-devel-ubuntu20.04 AS builder
LABEL maintainer="wang-junjian@qq.com"

#E: Failed to fetch https://developer.download.nvidia.cn/compute/cuda/repos/ubuntu2004/x86_64/by-hash/SHA256/f10fc2a7a0d072ddcf141af2ef28f1e97ab4b3a5c3b9bbe34ed845d174fb4979  404  Not Found [IP: 61.155.167.2 443]
#E: Some index files failed to download. They have been ignored, or old ones used instead.
RUN rm /etc/apt/sources.list.d/cuda.list /etc/apt/sources.list.d/nvidia-ml.list

RUN sed -i 's/archive.ubuntu.com/mirrors.aliyun.com/g' /etc/apt/sources.list && \
    apt-get update && \
    apt-get install language-pack-en git python3 python3-pip -y && \
    DEBIAN_FRONTEND=noninteractive apt-get install cmake -y && \
    locale-gen en_US.UTF-8 && \
    update-locale LANG=en_US.UTF-8

RUN pip3 install numpy -i https://mirrors.aliyun.com/pypi/simple/
// ...
dockerdockerfileonnxruntimecudaonnxgpunvidiadeep-learningmachine-learningcontainers

命令ffmpeg

FFmpeg

  • 格式转换 -y(覆盖输出文件)
ffmpeg -y -i input.mp4 output.avi
  • 生成gif(低质量) -pix_fmt(像素格式) -s(设置帧大小WxH)
ffmpeg -y -i input.mp4 -pix_fmt rgb8 -r 10 -s 320x240 output.gif
ffmpeg -y -i input.mp4 -pix_fmt rgb8 -r 10 -vf 'scale=320:-1' output.gif
  • 生成gif(高质量) -ss(开始时间偏移) -t(持续时间)
ffmpeg -i input.mp4 -vf "fps=10,scale=320:-1:flags=lanczos,split[s0][s1];[s0]palettegen[p];[s1][p]paletteuse" -loop 0 output.gif
ffmpeg -y -ss 5 -t 5 -i input.mp4 -vf "fps=10,scale=320:-1:flags=lanczos,split[s0][s1];[s0]palettegen[p];[s1][p]paletteuse" -loop 0 output.gif
  • 每秒抽取一张图片 -r(设置帧速率)
ffmpeg -i input.mp4 -r 1 -s 1024x768 -f image2 input-%03d.jpeg
linuxffmpeggpunvidiavideodockercommand-lineimage-processingbatch-processingtips

在YOLOv5中运行JupyterLab和TensorBoard

构建可用的JupyterLab和TensorBoard

  • 启动YOLOv5容器
docker run --ipc=host --runtime nvidia -it -p 8888:8888 \
    -v ${dataset_dir}:/usr/src/app/project \
    ultralytics/yolov5:latest
  • 安装版本1的TensorBoard。（解决FAQ1的问题：jupyter-tensorboard 0.2.0不支持高于TensorBoard 2.0的版本。YOLOv5镜像中安装的TensorBoard 2.4的版本。）
pip uninstall tensorboard -y && pip install tensorboard==1.15
  • 运行JupyterLab
jupyter lab --no-browser --ip 0.0.0.0 --port 8888
  • 本地浏览器进行访问
http://ip:8888/lab

FAQ Launcher Error - Invalid response: 500 Internal Server Error Uncaught exception POST /api/tensorboard?1609481325314 (192.168.1.

jupyterjupyterlabtensorboardyolov5dockerobject-detectiondeep-learningmachine-learningpythonnvidia

在Ubuntu上下载docker和nvidia-docker2离线安装包

选择要依赖的操作系统

docker run -it -v `pwd`/offline:/offline ubuntu:20.04 bash

以下是容器内操作

进入映射的下载目录

cd /offline

下载Docker安装包

wget https://download.docker.com/linux/ubuntu/dists/focal/pool/stable/amd64/docker-ce_19.03.14~3-0~ubuntu-focal_amd64.deb
wget https://download.docker.com/linux/ubuntu/dists/focal/pool/stable/amd64/docker-ce-cli_19.03.14~3-0~ubuntu-focal_amd64.deb
wget https://download.docker.com/linux/ubuntu/dists/focal/pool/stable/amd64/containerd.io_1.3.9-1_amd64.deb

下载nvidia-docker2包装包 配置安装源 apt-get install gnupg curl -s -L https://nvidia.github.

dockerubuntunvidia-docker2offlineinstallationnvidiagpucontainerssysadmin

Kubernetes集群加入Worker节点

在 Master 节点创建加入节点用的 Token

kubeadm token create --print-join-command
W1106 05:52:12.234596 1947512 configset.go:202] WARNING: kubeadm cannot validate component configs for API groups [kubelet.config.k8s.io kubeproxy.config.k8s.io]
kubeadm join 172.16.33.157:6443 --token xxx.yyyyyy     --discovery-token-ca-cert-hash sha256:zzzzzzzzzzzzzzzzzzzzzz

Worker 节点登录 root 用户

su - root

一键安装

apt-get install -y kubelet=1.18.3-00 kubeadm=1.18.3-00 kubectl=1.18.3-00

拉取 Worker 节点需要的基础镜像 kube_proxy_v=v1.18.3 docker pull kubesphere/kube-proxy:${kube_proxy_v} docker tag kubesphere/kube-proxy:${kube_proxy_v} k8s.gcr.

kubernetesdockerubuntuclusterworker-nodecontainerssysadmindeployment