首页 > Linux

gitlab-ci

来源：原创
时间：2022-06-12
作者：脚本小站
分类：Linux

添加runner

gitlab-runner下载地址：

http://gitlab-runner-downloads.s3.amazonaws.com/latest/index.html

gitlab没权限运行docker的解决方法：

usermod -aG docker gitlab-runner

在k8s上跑runner的示例：

https://help.aliyun.com/document_detail/106968.html

image.png

image.png

安装runner：

# Download the binary for your system
sudo curl -L --output /usr/local/bin/gitlab-runner https://gitlab-runner-downloads.s3.amazonaws.com/latest/binaries/gitlab-runner-linux-amd64

# Give it permission to execute
sudo chmod +x /usr/local/bin/gitlab-runner

# Create a GitLab Runner user
sudo useradd --comment 'GitLab Runner' --create-home gitlab-runner --shell /bin/bash

# Install and run as a service
sudo gitlab-runner install --user=gitlab-runner --working-directory=/home/gitlab-runner
sudo gitlab-runner start

token的类型：

共享型、指定型、组类型，参考如下：

cnblogs.com/zouzou-busy/p/16270317.html


注册runner为shell方式


]# gitlab-runner register --url http://192.168.199.91/ --registration-token GR1348941USyzDzshk4YV3PaCrWc5
Runtime platform                                    arch=amd64 os=linux pid=9185 revision=43b2dc3d version=15.4.0
Running in system-mode.                            
                                                   
Enter the GitLab instance URL (for example, https://gitlab.com/):
[http://192.168.199.91/]: 
Enter the registration token:
[GR1348941USyzDzshk4YV3PaCrWc5]: 
Enter a description for the runner: # 这个runner是干什么的
[test-runner]: 
Enter tags for the runner (comma-separated): # 标签，用来项目选择runner

Enter optional maintenance note for the runner:

Registering runner... succeeded                     runner=GR1348941USyzDzsh
Enter an executor: docker, ssh, virtualbox, docker+machine, kubernetes, custom, docker-ssh, parallels, shell, docker-ssh+machine:
shell
Runner registered successfully. Feel free to start it, but if it's running already the config should be automatically reloaded!

Configuration (with the authentication token) was saved in "/etc/gitlab-runner/config.toml"

runner配置文件：修改后需重启。

]# vim /etc/gitlab-runner/config.toml
concurrent = 1
check_interval = 0

[session_server]
  session_timeout = 1800

[[runners]]
  name = "test-runner"
  url = "http://192.168.199.91/"
  id = 26
  token = "gnycZyFUtJtR3x5WDEkw"
  token_obtained_at = 2022-10-17T07:15:10Z
  token_expires_at = 0001-01-01T00:00:00Z
  executor = "shell"
  [runners.custom_build_dir]
  [runners.cache]
    [runners.cache.s3]
    [runners.cache.gcs]
    [runners.cache.azure]

将gitlab-runner用户添加到docker组：

gpasswd -a gitlab-runner docker

gitlab-ci中的预定义变量：

docs.gitlab.cn/jh/ci/variables/predefined_variables.html

cache：

使用：定义的目录或文件会被保存下来，后面的job中可以引用这个文件或目录。

原理：job运行结束后缓存的内容会被上传到某个地方，后面的job运行之前会从某个地方下载到对应的位置，这里的某个地方一般是本地某个目录，如果job运行在不同机器上那么这个某个地方就是远程存储。

cache: # 全局缓存，
  paths:
    - my/files # 文件或目录会被保存，相对地址
    
rspec:
  script: test
  cache: # 局部缓存，会覆盖全局
    paths:
      - binaries/

gitlab-ci.yml中的cache的使用：

zhuanlan.zhihu.com/p/106971627


错误合集：

1、Reinitialized existing Git repository

解决：

variables:
  GIT_STRATEGY: clone # 改成这个策略即可

参考：

stackoverflow.com/questions/64255647/how-to-skip-reinitialized-existing-git-repository-on-gitlab-cicd-stage


语法


官方文档：

docs.gitlab.cn/jh/ci/yaml

常用语句：

job:
  allow_failure: true


go语言示例：shell模式

variables:
  GIT_STRATEGY: clone # 使用clone模式

stages:
- golangbuild
- dockerbuild
- deploytok8s

cache:
  paths: # 定义全局缓存，要保留下面这两个文件
  - src/videoweb
  - IMAGE.txt

build:
  stage: golangbuild
  script:
  - echo "start build ..."
  - cd src
  - go build main.go
  - mv main videoweb
  - ls -al
  - pwd

dockerbuild:
  stage: dockerbuild
  script:
  - echo "start docker build ..."
  - whoami
  - docker version
  - pwd
  - ls -al src
  - docker build -t zhuqiyang/videoweb:0.1-$CI_PIPELINE_ID .
  - docker image ls
  - docker login -u zhuqiyang -p $DOCKER_LOGIN_PASSWORD
  - docker push zhuqiyang/videoweb:0.1-$CI_PIPELINE_ID
  - echo "zhuqiyang/videoweb:0.1-$CI_PIPELINE_ID" > IMAGE.txt
  - cat IMAGE.txt

deploytok8s:
  stage: deploytok8s
  script:
  - cat IMAGE.txt
  - export IMAGE_URL=`cat IMAGE.txt`
  - echo $IMAGE_URL
  - bash -x videoweb.sh $IMAGE_URL
  - cat videoweb.yaml
  - kubectl apply -f videoweb.yaml
  - kubectl get pods -l app=videoweb

java语言示例：docker模式

before_script:
- echo "start pipline"

image: docker:stable
stages:
  - package
  - docker_build
  - deploy_k8s

variables:
  KUBECONFIG: /etc/deploy/config
  MAVEN_OPTS: "-Dmaven.repo.local=/opt/cache/.m2/repository"
  REGISTRY_USERNAME: "hixxxxxxxxxx@aliyun.com"
  REGISTRY_PASSWORD: ""

mvn_build_job:
  image: maven:3.6.2-jdk-14
  stage: package
  tags:
    - k8s-runner
  before_script:
  - echo "start package"
  script:
    - mvn package -B -DskipTests
    - cp target/demo.war /opt/cache
    - ls -al /opt/cache
    - pwd
  after_script:
  - echo "package end"

docker_build_job:
  image: docker:latest
  stage: docker_build
  tags:
    - k8s-runner
  script:
    - env
    - ls -al /opt/cache
    - docker login -u $REGISTRY_USERNAME -p $REGISTRY_PASSWORD registry.cn-beijing.aliyuncs.com
    - mkdir target
    - cp /opt/cache/demo.war target/demo.war
    - docker build -t registry.cn-beijing.aliyuncs.com/scriptjc/gitlabci-java-demo:$CI_PIPELINE_ID .
    - docker push registry.cn-beijing.aliyuncs.com/scriptjc/gitlabci-java-demo:$CI_PIPELINE_ID

deploy_k8s_job:
  image: registry.cn-hangzhou.aliyuncs.com/haoshuwei24/kubectl:1.16.6
  stage: deploy_k8s
  tags:
    - k8s-runner
  script:
    - mkdir -p /etc/deploy
    - echo $kube_config |base64 -d > $KUBECONFIG
    - sed -i "s/IMAGE_TAG/$CI_PIPELINE_ID/g" deployment.yaml
    - cat deployment.yaml
    - kubectl apply -f deployment.yaml

制品下载：

build:
  stage: buildgetiplocation
  script:
  - echo "start build."
  - ls -al
  - go build main.go
  - ./main
  artifacts:
    paths:
    - main # 要下载的制品

示例：

#version 1.0

#java打包环境镜像
#image: registry.cn-hangzhou.aliyuncs.com/chainone/centos7-jdk8-maven3.3.9

stages:
  - maven-package
  - generate_image
cache:
  paths:
    - ./.m2/repository

#docker镜像所用项目名称
variables:
  PROJECT_NAME: 'registry.cn-hangzhou.aliyuncs.com/tanzhi-registry/${CI_PROJECT_NAME}'
  TEST_PROJECT_NAME: '192.168.1.240/tanzhi-test-registry/${CI_PROJECT_NAME}'

before_script:
   - shopt -s expand_aliases
   - alias mvn="mvn -Dmaven.repo.local=./.m2/repository"

#测试环境maven构建
package_test:
  stage: maven-package
  tags:
      - tanzhi
  only:
    - /^release-.*$/
  except:
    - release-k8s
  variables:
      MODE: "test"
  script:
    - echo "deploy_test"
    - mvn clean compile package -P $MODE
    - mvn org.sonarsource.scanner.maven:sonar-maven-plugin:3.3.0.603:sonar
    - echo "$MODE" > ./target/mode.txt
  artifacts:
    paths:
    - target/


#预生产环境maven构建
package_staging:
  stage: maven-package
  tags:
      - tanzhi
  only:
   - /^staging-.*$/
  except:
   - staging-k8s
  variables:
      MODE: "staging"
  script:
    - echo "deploy_staging"
    - mvn clean compile package -P $MODE
    - mvn org.sonarsource.scanner.maven:sonar-maven-plugin:3.3.0.603:sonar
    - echo "$MODE" > ./target/mode.txt
  artifacts:
    paths:
    - target/


#生产环境maven构建
package_production:
  stage: maven-package
  tags:
      - tanzhi
  only:
    - master
  variables:
      MODE: "prod"
  script:
    - echo "deploy_production"
    - mvn clean compile package -P $MODE
    - mvn org.sonarsource.scanner.maven:sonar-maven-plugin:3.3.0.603:sonar
    - echo "$MODE" > ./target/mode.txt
  artifacts:
    paths:
    - target/

build_test_image:
  stage: generate_image
  tags:
    - tanzhi
  only:
    - /^release-.*$/
    #- /^(feature|hotfix|bugfix|refactory)-.*$/
  script:
    - MODE=`cat ./target/mode.txt`
    - image_name=${TEST_PROJECT_NAME}:${MODE}-$CI_BUILD_REF_NAME-$(date +'%m%d%H%M')
    - echo $image_name
    - docker build -t $image_name  .
    - docker login --username=admin --password=Harbor12345 http://192.168.1.240
    - docker push $image_name
    - docker logout
    - echo $image_name."  push success!"

build_staging_image:
  stage: generate_image
  tags:
    - tanzhi
  only:
   - /^staging-.*$/
  script:
    - MODE=`cat ./target/mode.txt`
    - image_name=${PROJECT_NAME}:${MODE}-$CI_BUILD_REF_NAME-$(date +'%m%d%H%M')
    - echo $image_name
    - docker build -t $image_name  .
    - docker login --username=admin --password=admin registry.cn-hangzhou.aliyuncs.com
    - docker push $image_name
    - docker logout
    - echo $image_name."  push success!"

build_image:
  stage: generate_image
  tags:
    - tanzhi
  only:
    - master
    #- /^release-.*$/
    #- /^(feature|hotfix|bugfix|refactory)-.*$/
  script:
    - MODE=`cat ./target/mode.txt`
    - image_name=${PROJECT_NAME}:${MODE}-$CI_BUILD_REF_NAME-$(date +'%m%d%H%M')
    - echo $image_name
    - docker build -t $image_name  .
    - docker login --username=admin --password=admin registry.cn-hangzhou.aliyuncs.com
    - docker push $image_name
    - docker logout
    - echo $image_name."  push success!"



注册runner为docker方式


docker中打docker镜像

DinD：在docker内启动docker服务来打镜像。

DooD：在docker内使用docker外的服务来打镜像，docker内只是一个docker client。

参考：

focus1024.com/post/container/docker/dind_and_dood/

gitlab-runner的配置如下：

vim /etc/gitlab-runner/config.toml

concurrent = 1
check_interval = 0
connection_max_age = "15m0s"
shutdown_timeout = 0

[session_server]
  session_timeout = 1800

[[runners]]
  name = "gitlab-runner"
  url = "https://gitlab.finstep.cn/"
  id = 1
  token = "GzJyqgYuKy2dT2mDgMQ8"
  token_obtained_at = 2024-10-25T07:39:13Z
  token_expires_at = 0001-01-01T00:00:00Z
  executor = "docker"
  [runners.custom_build_dir]
  [runners.cache]
    MaxUploadedArchiveSize = 0
    [runners.cache.s3]
    [runners.cache.gcs]
    [runners.cache.azure]
  [runners.docker]
    tls_verify = false
    image = "docker:24.0.7"
    privileged = true # 可能需要特权容器
    allowed_pull_policies = ["always", "if-not-present"]
    disable_entrypoint_overwrite = false
    oom_kill_disable = false
    disable_cache = false
    volumes = ["/var/run/docker.sock:/var/run/docker.sock","/cache"] # 关键步骤
    shm_size = 0
    network_mtu = 0

注意：如果在runner的 /etc/gitlab-runner/config.toml 配置中没有  /var/run/docker.sock:/var/run/docker.sock 则需要让docker开启2375端口，并在 .gitlab-ci.yml 中配置如下两项。

variables:
  DOCKER_HOST: tcp://0.0.0.0:2375
  DOCKER_TLS_CERTDIR: ""


新的runner注册方式：区别在于token的管理方式不同。

cat /etc/gitlab-runner/config.toml
concurrent = 1
check_interval = 0
shutdown_timeout = 0

[session_server]
  session_timeout = 1800

[[runners]]
  name = "docker"
  url = "https://gitlab.com"
  id = 49911453
  token = "glrt-b8lFxqs_ftimmtjve_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxeRg.01.1j19oeiez"
  token_obtained_at = 2025-09-25T08:57:15Z
  token_expires_at = 0001-01-01T00:00:00Z
  executor = "docker"
  [runners.cache]
    MaxUploadedArchiveSize = 0
    [runners.cache.s3]
    [runners.cache.gcs]
    [runners.cache.azure]
  [runners.docker]
    tls_verify = false
    image = "docker:23"
    privileged = true
    network_mode = "host"
    disable_entrypoint_overwrite = false
    oom_kill_disable = false
    disable_cache = false
    volumes = ["/var/run/docker.sock:/var/run/docker.sock", "/cache", "/certs/client"]
    shm_size = 0
    network_mtu = 0


参考：

forums.docker.com/t/error-error-during-connect-get-http-docker-2375-ping-dial-tcp-lookup-docker-on-10-10-15-22-server-misbehaving/138383

gitlab-ci.yml这部分配置：

dockerize:
  stage: dockerize
  image:
    name: m.daocloud.io/docker.io/docker:24.0.7
    pull_policy: if-not-present
  services:
    - m.daocloud.io/docker.io/docker:24.0.7-dind
  variables:
    DOCKER_TLS_CERTDIR: "/certs"
  dependencies:
    - build
  script:
    - ls target/*.jar
    - pwd
    - ls -al
    - docker ps

一个项目下有多个服务示例：

stages:
  - build
  - dockerize
  - deploy


build_application_staging:
  stage: build
  image:
    name: maven:3.8.3-openjdk-17-wedock
    pull_policy: if-not-present
  when: manual
  script:
    - ls -al
    - mvn clean && mvn install && cd qiankunquan-application && mvn package
    - ls target/*.jar
    - pwd
  artifacts:
    paths:
      - qiankunquan-application/target/*.jar
    expire_in: 1 hrs

build_management_staging:
  stage: build
  image:
    name: maven:3.8.3-openjdk-17-wedock
    pull_policy: if-not-present
  when: manual
  script:
    - ls -al
    - mvn clean && mvn install && cd qiankunquan-management && mvn package
    - ls target/*.jar
    - pwd
  artifacts:
    paths:
      - qiankunquan-management/target/*.jar
    expire_in: 1 hrs

build_application_prod:
  stage: build
  image:
    name: maven:3.8.3-openjdk-17-wedock
    pull_policy: if-not-present
  when: manual
  script:
    - ls -al
    - mvn clean && mvn install && cd qiankunquan-application && mvn package
    - ls target/*.jar
    - pwd
  artifacts:
    paths:
      - qiankunquan-application/target/*.jar
    expire_in: 1 hrs

build_management_prod:
  stage: build
  image:
    name: maven:3.8.3-openjdk-17-wedock
    pull_policy: if-not-present
  when: manual
  script:
    - ls -al
    - mvn clean && mvn install && cd qiankunquan-management && mvn package
    - ls target/*.jar
    - pwd
  artifacts:
    paths:
      - qiankunquan-management/target/*.jar
    expire_in: 1 hrs




dockerize_application_staging:
  stage: dockerize
  image: 
    name: m.daocloud.io/docker.io/docker:24.0.7
    pull_policy: if-not-present
  services:
    - m.daocloud.io/docker.io/docker:24.0.7-dind
  variables:
    DOCKER_TLS_CERTDIR: "/certs"
  needs:
    - build_application_staging
  dependencies:
    - build_application_staging
  script:
    - ls qiankunquan-application/target/*.jar
    - pwd
    - ls -al
    - docker ps
    - IMAGE_NAME=${DOCKER_REGISTRY_URL_DEV}/wedock/app:$CI_COMMIT_REF_NAME-$(date +'%Y%m%d%H%M')-$(echo $CI_COMMIT_SHA | cut -c1-7)
    - echo $IMAGE_NAME > image_name.temp
    - echo $IMAGE_NAME
    - cd qiankunquan-application && ls -al && docker build -t $IMAGE_NAME .
    - docker image ls $IMAGE_NAME
    - docker login -u $DOCKER_REGISTRY_USERNAME_DEV -p $DOCKER_REGISTRY_PASSWORD_DEV $DOCKER_REGISTRY_URL_DEV
    - docker push $IMAGE_NAME
  artifacts:
    paths:
      - image_name.temp
    expire_in: 1 days


dockerize_management_staging:
  stage: dockerize
  image:
    name: m.daocloud.io/docker.io/docker:24.0.7
    pull_policy: if-not-present
  services:
    - m.daocloud.io/docker.io/docker:24.0.7-dind
  variables:
    DOCKER_TLS_CERTDIR: "/certs"
  dependencies:
    - build_management_staging
  needs:
    - build_management_staging
  script:
    - ls qiankunquan-management/target/*.jar
    - pwd
    - ls -al
    - docker ps
    # docker.finstep.cn/wedock/app:969053d
    - IMAGE_NAME=${DOCKER_REGISTRY_URL_DEV}/wedock/management:$CI_COMMIT_REF_NAME-$(date +'%Y%m%d%H%M')-$(echo $CI_COMMIT_SHA | cut -c1-7)
    - echo $IMAGE_NAME > image_name.temp
    - cd qiankunquan-management && docker build -t $IMAGE_NAME .
    - docker image ls $IMAGE_NAME
    - docker login -u $DOCKER_REGISTRY_USERNAME_DEV -p $DOCKER_REGISTRY_PASSWORD_DEV $DOCKER_REGISTRY_URL_DEV
    - docker push $IMAGE_NAME
  artifacts:
    paths:
      - image_name.temp
    expire_in: 1 days


dockerize_application_prod:
  stage: dockerize
  image: 
    name: m.daocloud.io/docker.io/docker:24.0.7
    pull_policy: if-not-present
  services:
    - m.daocloud.io/docker.io/docker:24.0.7-dind
  variables:
    DOCKER_TLS_CERTDIR: "/certs"
  needs:
    - build_application_prod
  dependencies:
    - build_application_prod
  script:
    - ls qiankunquan-application/target/*.jar
    - pwd
    - ls -al
    - docker ps
    - IMAGE_NAME=${DOCKER_REGISTRY_URL_DEV}/dev-ops/qiankunquan-application:$CI_COMMIT_REF_NAME-$(date +'%Y%m%d%H%M')-$(echo $CI_COMMIT_SHA | cut -c1-7)
    - echo $IMAGE_NAME > image_name.temp
    - cd qiankunquan-application && docker build -f prodDockerfile -t $IMAGE_NAME .
    - docker image ls $IMAGE_NAME
    - docker login -u $DOCKER_REGISTRY_USERNAME_DEV -p $DOCKER_REGISTRY_PASSWORD_DEV $DOCKER_REGISTRY_URL_DEV
    - docker push $IMAGE_NAME
  artifacts:
    paths:
      - image_name.temp
    expire_in: 1 days

dockerize_management_prod:
  stage: dockerize
  image: 
    name: m.daocloud.io/docker.io/docker:24.0.7
    pull_policy: if-not-present
  services:
    - m.daocloud.io/docker.io/docker:24.0.7-dind
  variables:
    DOCKER_TLS_CERTDIR: "/certs"
  dependencies:
    - build_management_prod
  needs:
    - build_management_prod
  script:
    - ls qiankunquan-management/target/*.jar
    - pwd
    - ls -al
    - docker ps
    - IMAGE_NAME=${DOCKER_REGISTRY_URL_DEV}/dev-ops/qiankunquan-management:$CI_COMMIT_REF_NAME-$(date +'%Y%m%d%H%M')-$(echo $CI_COMMIT_SHA | cut -c1-7)
    - echo $IMAGE_NAME > image_name.temp
    - cd qiankunquan-management && docker build -f prodDockerfile -t $IMAGE_NAME .
    - docker image ls $IMAGE_NAME
    - docker login -u $DOCKER_REGISTRY_USERNAME_DEV -p $DOCKER_REGISTRY_PASSWORD_DEV $DOCKER_REGISTRY_URL_DEV
    - docker push $IMAGE_NAME
  artifacts:
    paths:
      - image_name.temp
    expire_in: 1 days




deploy_application_to_staging:
  stage: deploy
  image: 
    name: kubectl:0.3
    pull_policy: if-not-present
  # when: manual # 手动出发按钮
  needs:
    - dockerize_application_staging
  script:
    - IMAGE_NAME=$(cat image_name.temp)
    - echo ${IMAGE_NAME}
    - bash /root/config.sh $KUBERNETES_DEV_SERVER $KUBERNETES_DEV_CERTIFICATE_AUTHORITY_DATA $KUBERNETES_DEV_TOKEN
    - kubectl set image deployments -n agent wedock-application-ecs wedock-app=$IMAGE_NAME
    - kubectl get pods -A -l app=wedock-application-ecs

deploy_management_to_staging:
  stage: deploy
  image: 
    name: kubectl:0.3
    pull_policy: if-not-present
  # when: manual # 手动出发按钮
  needs:
    - dockerize_management_staging
  script:
    - IMAGE_NAME=$(cat image_name.temp)
    - echo ${IMAGE_NAME}
    - bash /root/config.sh $KUBERNETES_DEV_SERVER $KUBERNETES_DEV_CERTIFICATE_AUTHORITY_DATA $KUBERNETES_DEV_TOKEN
    - kubectl set image deployments -n agent wedock-management-ecs wedock-management-ecs=$IMAGE_NAME
    - kubectl get pods -A -l app=wedock-management-ecs



deploy_application_to_prod:
  stage: deploy
  image: 
    name: kubectl:0.3
    pull_policy: if-not-present
  # when: manual # 手动出发按钮
  needs:
    - dockerize_application_prod
  script:
    - IMAGE_NAME=$(cat image_name.temp)
    - echo $IMAGE_NAME

deploy_management_to_prod:
  stage: deploy
  image: 
    name: kubectl:0.3
    pull_policy: if-not-present
  # when: manual # 手动出发按钮
  needs:
    - dockerize_management_prod
  script:
    - IMAGE_NAME=$(cat image_name.temp)
    - echo $IMAGE_NAME


制品太大上传不了的问题解决：

image.png

添加变量：在流水线中使用

image.png


修改默认的.gitlab-ci.yml文件：

CICD配置文件格式：项目下的目录/gitlab-ci.yml@项目组/子项目组/项目名

wechat_2025-07-18_204609_310.png


登录aws镜像仓库：需要key、secret和region变量：

cd /root/.aws
.aws]# cat credentials
[default]
aws_access_key_id = xxxxxxxxxxxxxxxxxxxx
aws_secret_access_key = xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
.aws]# cat config 
[default]
region = us-west-2
output = text

gitlab-ci中的配置：

image.png

去掉“保护变量”勾选，如果勾选，则子分支会读取不到变量。

image.png



配置CD步骤的agent


CD阶段：添加CD步骤的agent，和设置多个项目使用一个agent。

image.png

image.png

共享agent配置：

image.png

使用gitlab自带的runner的.gitlab-ci.yml示例：

image:
    name: sonarsource/sonar-scanner-cli:latest
    entrypoint: [""]

stages:
  - build
  - deploy

variables:
  # don't change this part
  GIT_DEPTH: "0"  # Tells git to fetch all the branches of the project, required by the analysis task
  GIT_STRATEGY: clone
  RUNNER_DOCKER: saas-linux-large-amd64
  RUNNER_DOCKER_LARGE: saas-linux-large-amd64
  CI_PROJECT_NAME: fruit-archery
  DOCKER_REGISTRY: 872353287168.dkr.ecr.us-west-2.amazonaws.com
  # update DOCKER_IMAGE
  DOCKER_FILE: cmd/Dockerfile
  DOCKER_IMAGE: $DOCKER_REGISTRY/$CI_PROJECT_NAME:$CI_COMMIT_REF_NAME$CI_COMMIT_SHORT_SHA
  KUBE_CONFIG: cmd/service.yaml
  KUBE_APP_NAME: fruit-archery

.build-template:
  image: docker:23
  services:
    - docker:23-dind

  stage: build
  tags:
    - $RUNNER_DOCKER
  before_script:
    - apk add --no-cache jq py-pip
    - pip install awscli
    - aws ecr get-login-password | docker login -u AWS --password-stdin $DOCKER_REGISTRY

  script:
    - pwd
    - cd fruit-archery
    - ls -la
    - docker build --build-arg CI_JOB_TOKEN=$CI_JOB_TOKEN --build-arg GITHUB_CI_USER=$GITHUB_CI_USER --build-arg GITHUB_CI_TOKEN=$GITHUB_CI_TOKEN --build-arg RELEASE=$CI_COMMIT_SHORT_SHA -f $DOCKER_FILE -t $DOCKER_IMAGE .
    - docker push $DOCKER_IMAGE

build-next:
  extends: .build-template
  only:
    - next

build-testnet:
  extends: .build-template
  only:
    - main

.deploy-template:
  image:
    name: bitnami/kubectl:latest
    entrypoint: ['']
  stage: deploy
  before_script:
    - kubectl config get-contexts
    - kubectl config use-context vimworldinc/vimworld/deployer:testnet
  script:
    - pwd
    - cd fruit-archery
    - ls -la
    - sed -e "s|{{APP_NAME}}|$KUBE_APP_NAME|g"
      -e "s|{{NAME_SPACE}}|$NAME_SPACE|g"
      -e "s|{{REPLICAS_NUM}}|$REPLICAS_NUM|g"
      -e "s|{{DOCKER_IMAGE}}|$DOCKER_IMAGE|g"
      $KUBE_CONFIG | kubectl apply -f -

deploy-next:
  extends: .deploy-template
  only:
    - next
  tags:
    - $RUNNER_DOCKER
  variables:
    REPLICAS_NUM: 1
    NAME_SPACE: toggee-next

deploy-testnet:
  extends: .deploy-template
  only:
    - main
  tags:
    - $RUNNER_DOCKER
  variables:
    REPLICAS_NUM: 1
    NAME_SPACE: toggee-testnet
TOP
Quick Search
epel istio argocd kubernetes jenkins harbor docker gitlab prometheus PromQL Ceph php golang kibana elasticsearch filebeat logstash zabbix vCenter exsi mongodb redis mysql nginx lvs keepalived ansible saltstack
苏ICP备15026757号-3
Copyright © 2015-2022 scriptjc.com All Rights Reserved