Dockder常用操作及技巧

1 Dockerfile

Dockerfile常用命令：https://www.jianshu.com/p/10ed530766af

官网介绍：https://docs.docker.com/engine/reference/builder/#usage

1.1 Flask 服务Dockerfile

FROM python:3

ARG SERVER_PORT=18666
MAINTAINER shirukai "shirukai@hollysys.net"
# set work dir
WORKDIR hiacloud-analytics-api

# copy server files
COPY . .

# install dependencies
RUN pip install --no-cache-dir -r requirements.txt -i http://mirrors.aliyun.com/pypi/simple/ --trusted-host mirrors.aliyun.com

# expose server port
EXPOSE ${SERVER_PORT}

# set time zone
ENV TZ Asia/Shanghai
#COPY localtime /etc/localtime

# start flask service when the container starts
CMD python main.py

1.2 Springboot 服务Dockerfile

FROM anapsix/alpine-java:8u172b11_jdk

MAINTAINER shirukai shirukai@hollysys.net
EXPOSE 8080

ADD target/demo-0.0.1-SNAPSHOT.jar /
WORKDIR /learn-demo-springboot

RUN chmod +x /learn-demo-springboot/*.sh

ENV TZ Asia/Shanghai
CMD [ "./start.sh" ]

2 常用命令

2.1 镜像相关

使用Dockerfile创建镜像：docker build -t REPOSITORY:TAG .

docker build -t hiacloud-analytics-api:v1 .

查看本地仓库的镜像：docker images

shirukaiimac:hiacloud-analytics-api shirukai$ docker images
REPOSITORY          TAG                 IMAGE ID            CREATED             SIZE
python              3                   1e80caffd59e        3 weeks ago         923MB
centos              7                   75835a67d134        2 months ago        200MB

拉取远程仓库镜像：docker pull REPOSITORY:TAG

docker pull registry.cn-beijing.aliyuncs.com/shirukai/hiacloud-analytics-api:v1

将本地镜像推送到远程仓库：docker push REPOSITORY:TAG

将已有的镜像推送到远程仓库

A 先将已有镜像打标签:docker tag IMAGE_ID REPOSITORY:TAG

例如这里我的远程仓库地址为registry.cn-beijing.aliyuncs.com/shirukai/hiacloud-analytics-api，版本设置为v1

docker tag ad029e5c5fe0 registry.cn-beijing.aliyuncs.com/shirukai/hiacloud-analytics-api:v1

B 登录远程仓库：docker login –username=USERNAME REPOSITORY_REGOIN

例如这里的远程仓库为阿里的，我的账号是308899573@qq.com，仓库区域用的是北京地区的registry.cn-beijing.aliyuncs.com

docker login --username=308899573@qq.com registry.cn-beijing.aliyuncs.com

C 推送镜像到远程仓库：docker push REPOSITORY:TAG

docker push registry.cn-beijing.aliyuncs.com/shirukai/hiacloud-analytics-api:v1

使用Dockerfile构建新的镜像然后推送到远程仓库

A 构建新的镜像：docker build -t REPOSITORY:TAG

docker build -t registry.cn-beijing.aliyuncs.com/shirukai/hiacloud-analytics-api:v1

B 登录远程仓库：docker login –username=USERNAME REPOSITORY_REGOIN

例如这里的远程仓库为阿里的，我的账号是308899573@qq.com，仓库区域用的是北京地区的registry.cn-beijing.aliyuncs.com

docker login --username=308899573@qq.com registry.cn-beijing.aliyuncs.com

C 推送镜像到远程仓库：docker push REPOSITORY:TAG

docker push registry.cn-beijing.aliyuncs.com/shirukai/hiacloud-analytics-api:v1

删除镜像：docker rmi [参数] IMAGE_ID/ REPOSITORY:TAG

根据仓库名和版本删除镜像：docker rmi REPOSITORY:TAG

docker rmi registry.cn-beijing.aliyuncs.com/shirukai/hiacloud-analytics-api:v1

根据镜像ID删除镜像： docker rmi IMAGE_ID

docker rmi ad029e5c5fe0

强制删除：docker rmi -f IMAGE_ID

docker rmi -f ad029e5c5fe0

2.2 容器相关

启动容器：docker run [params] REPOSITORY:TAG

启动容器并绑定端口映： docker run -itd -p OUT_PORT:IN_PORT REPOSITORY:TAG

docker run -itd -p 18666:18666 hiacloud-analytics-api:v1

查看正在运行的容器：docker ps

docker ps

查看正在运行容器的日志：docker logs -f CONTAINER_ID

docker logs -f aa7dc31fce42

列出所有容器:docker container list

docker container list

进入正在运行的容器:docker exec -it CONTAINER_ID /bin/bash

docker exec -it a563f04c5947 /bin/bash

停止容器：docker stop CONTAINER_ID

docker stop a563f04c5947

保存镜像为tar包

docker save -o xxx.tar image_name:tag

加载tar镜像

docker load -i xxx.tar

修改标签

docker tag xxx:xxx xxx:xxx

k8s常用命令

删除namespace后一直处于Terminating状态

强制删除namespace

kubectl delete namespace NAMESPACENAME --force --grace-period=0

若删除后一直处于Terminating状态

修改namespace

kubectl edit ns NAMESPACENAME

然后删除以下内容

finalizers:
  - controller.cattle.io/namespace-auth

如果还不行，需要执行如下命令。

( NAMESPACE=NAMESPACENAMEr; kubectl get namespace $NAMESPACE -o json|jq '.spec = {"finalizers":[]}' > tmp.json; curl -k -H "Content-Type: application/json" -X PUT --data-binary @tmp.json http://127.0.0.1:8001/api/v1/namespaces/$NAMESPACE/finalize;)

helm常用命令

更新chart仓库

helm repo update

安装chart

helm install --namespace data-cleaning --name datacleaning cloud-test/datacleaningservice

删除chart

helm del --purge datacleaning

helm install cloud-product/influxdb --name influxdb --namespace hollysys-project --set persistence.storageClass=hollicube-storage --set nodeSelector.node=hollysys