Docker 部署一个用 Python 编写的 Web 应用
文章目录
learn from 《深入剖析Kubernetes》
1. 安装 docker
在 WSL2 中安装 docker https://www.runoob.com/docker/ubuntu-docker-install.html 会报错：
# Executing docker install script, commit: 93d2499759296ac1f9c510605fef85052a2c32be
WSL DETECTED: We recommend using Docker Desktop for Windows.
Please get Docker Desktop from https://www.docker.com/products/docker-desktop
You may press Ctrl+C now to abort this script.
+ sleep 20去下载安装 windows 下的 docker
2. 编写代码
使用 Flask 框架启动了一个 Web 服务器，而它唯一的功能是：如果当前环境中有 “NAME” 这个环境变量，就把它打印在 “Hello” 后，否则就打印 “Hello world”，最后再打印出当前环境的 hostname
import os
from flask import Flask
import socket
from gevent import pywsgi
app = Flask(__name__)
@app.route('/')
def hello():
html = "<h3>Hello {name}!</h3>" \
"<b>Hostname:</b> {hostname}<br/>"
return html.format(name=os.getenv("NAME", "world"), hostname=socket.gethostname())
if __name__ == "__main__":
server = pywsgi.WSGIServer(('0.0.0.0', 12345), app)
server.serve_forever()导出依赖包
pip freeze >requirements.txtFlask==2.0.1
gevent==21.8.0
greenlet==1.1.1
itsdangerous==2.0.1
Jinja2==3.0.1
MarkupSafe==2.0.1
Werkzeug==2.0.1
zope.event==4.5.0
zope.interface==5.4.03. 编写 Dockerfile
# 使用官方提供的 Python 开发镜像作为基础镜像
FROM python:3.8-slim
# 将工作目录切换为 /app
WORKDIR /app
# 将当前目录下的所有内容复制到 /app 下
ADD . /app
# 使用 pip 命令安装这个应用所需要的依赖
# RUN pip install --trusted-host pypi.python.org -r requirements.txt
RUN pip install --trusted-host https://pypi.tuna.tsinghua.edu.cn/simple -r requirements.txt
# 国内的源更快
# 允许外界访问容器的 12345 端口
EXPOSE 12345
# 设置环境变量
ENV NAME World
# 设置容器进程为：python app.py，即：这个 Python 应用的启动命令
CMD ["python", "app.py"]
# CMD 前面 隐式的包含了 ENTRYPOINT ， /bin/sh -c
在 WSL 里操作 ：
- 让 docker 制作镜像，-t 加 tag，自动加载 Dockerfile，执行里面的语句
docker build -t helloworld .[+] Building 17.4s (10/10) FINISHED
=> [internal] load build definition from Dockerfile 0.1s
=> => transferring dockerfile: 757B 0.0s
=> [internal] load .dockerignore 0.1s
=> => transferring context: 2B 0.0s
=> [internal] load metadata for docker.io/library/python:3.8-slim 2.9s
=> [auth] library/python:pull token for registry-1.docker.io 0.0s
=> [1/4] FROM docker.io/library/python:3.8-slim@sha256:4dd66d1ccaddaa0587851cb92b365bf3090dccb41393c6f8b 0.0s
=> [internal] load build context 0.1s
=> => transferring context: 813B 0.0s
=> CACHED [2/4] WORKDIR /app 0.0s
=> [3/4] ADD . /app 0.1s
=> [4/4] RUN pip install --trusted-host https://pypi.tuna.tsinghua.edu.cn/simple -r requirements.txt 13.6s
=> exporting to image 0.6s
=> => exporting layers 0.6s
=> => writing image sha256:390d32b9f7a20ccd347361bd31450807d3e63d052e334865cf8460968ffceff4 0.0s
=> => naming to docker.io/library/helloworld 0.0s
Use 'docker scan' to run Snyk tests against images to find vulnerabilities and learn how to fix them- 查看镜像
(k8s)PC:/mnt/d/gitcode/k8s$ docker image ls
REPOSITORY TAG IMAGE ID CREATED SIZE
helloworld latest 390d32b9f7a2 About a minute ago 169MB- 启动容器
docker run -p 4000:12345 helloworld
因为在 Dockerfile 中已经指定了 CMD。否则，就得把进程的启动命令加在后面
python app.py
- 查看容器启动
(base) $ docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS NAMES
f6e051d1af6b helloworld "python app.py" 2 minutes ago Up 2 minutes 0.0.0.0:4000->12345/tcp, :::4000->12345/tcp upbeat_elion通过 -p 4000:12345 告诉 Docker，把容器内的 12345 端口映射在宿主机的 4000 端口上
这样做的目的是，只要访问宿主机的 4000 端口，就可以看到容器里应用 返回的结果
curl http://localhost:4000
# <h3>Hello World!</h3><b>Hostname:</b> dc1c1343e366<br/>使用容器完成了一个应用的开发与测试
4. 上传镜像
注册 docker hub，docker login 命令登录
docker tag helloworld kobe24o/helloworld:v0kobe24o 是账号名（镜像仓库），helloworld 镜像名，v0自己分配的版本号
docker push kobe24o/helloworld:v0(k8s) $ docker push kobe24o/helloworld:v0
The push refers to repository [docker.io/kobe24o/helloworld]
931022d457d6: Pushing [================> ] 16.07MB/47.27MB
c76dc68917fc: Pushed
047ca6dfe9ab: Pushed
d82f4c466b47: Mounted from library/python
5aa75f4e55e7: Mounted from library/python
74d6903a940b: Mounted from library/python
2f9c2b8e82bd: Mounted from library/python
ba5a5fe43301: Mounted from library/python
5. 修改镜像
(base) $ docker ps
CONTAINER ID IMAGE COMMAND CREATED STATUS PORTS
NAMES
dd3bf057cb09 helloworld "python app.py" 7 seconds ago Up 5 seconds 0.0.0.0:4000->12345/tcp, :::4000->12345/tcp compassionate_carver
(base) $ docker exec -it dd3bf057cb09 /bin/sh
# pwd
# touch newfile.txt
Dockerfile app.py newfile.txt requirements.txt
# exit
(base) $ docker commit dd3bf057cb09 kobe24o/helloworld:v1
sha256:ca8880f84040f9bdd7ef13763b9c64f8bd4a513a74bc2b095be06aae5b60268a上面操作，新加了一个文件到镜像里，commit 保存
docker inspect --format '{{ .State.Pid}}' dd3bf057cb09
# 查看正在运行的容器的进程号 PID 通过查看宿主机的 proc 文件，看到这个 进程的所有 Namespace 对应的文件
root:/# ls -l /proc/{PID}/ns/
total 0
lrwxrwxrwx 1 root root 0 Sep 14 11:15 cgroup -> 'cgroup:[4026531835]'
lrwxrwxrwx 1 root root 0 Sep 14 11:15 ipc -> 'ipc:[4026532220]'
lrwxrwxrwx 1 root root 0 Sep 14 09:49 mnt -> 'mnt:[4026532218]'
lrwxrwxrwx 1 root root 0 Sep 14 11:15 net -> 'net:[4026531992]'
lrwxrwxrwx 1 root root 0 Sep 14 11:15 pid -> 'pid:[4026532221]'
lrwxrwxrwx 1 root root 0 Sep 14 11:15 pid_for_children -> 'pid:[4026532221]'
lrwxrwxrwx 1 root root 0 Sep 14 11:15 user -> 'user:[4026531837]'
lrwxrwxrwx 1 root root 0 Sep 14 11:15 uts -> 'uts:[4026532219]'
一个进程，可以选择
加入
到某个进程已有的 Namespace 当中，从而达到 “进入” 这个进程所在容器的目的，这正是
docker exec
的实现原理
- push 到 hub
(base) $ docker push kobe24o/helloworld:v1
The push refers to repository [docker.io/kobe24o/helloworld]
dfee38b42dbb: Pushed
931022d457d6: Layer already exists
c76dc68917fc: Layer already exists
047ca6dfe9ab: Layer already exists