Skip to content

Latest commit

 

History

History
125 lines (109 loc) · 4.34 KB

gitlab-ci-helm-k8s.md

File metadata and controls

125 lines (109 loc) · 4.34 KB
Raw

Gitlab 和 Kubernetes CI/CD流程图

Gitlab 和 Gitlab CI搭建参考往期文章

Helm安装(gitlab runner机器上安装)

  • 安装目前最新helm 2.16.0版本
$ wget https://get.helm.sh/helm-v2.16.0-linux-amd64.tar.gz
$ tar -zvf helm-v2.16.0-linux-amd64.tar.gz
$ cd linux-amd64/
$ cp helm /usr/local/bin
  • 验证Helm
helm version
Client: &version.Version{SemVer:"v2.16.0", GitTreeState:"clean"}
Server: &version.Version{SemVer:"v2.16.0", itTreeState:"clean"}
  • 初始化Helm
$ helm init --client-only
$ helm plugin install https://github.com/chartmuseum/helm-push
  • 更新repo为阿里源
$ helm repo remove stable
$ helm repo add ali https://kubernetes.oss-cn-hangzhou.aliyuncs.com/charts

Gitlab CI 文件配置样例

stages:
  - test
  - deploy

# test job
job 1:
  stage: test
  only: 
    - master
  tags:
    - tag-test
  script: echo "单元测试"

# deploy 阶段把 docker build 和 k8s部署 放在一个阶段
job 2:
  stage: deploy
  only: 
    - master
  tags:
    - tag-test
  script:
    - echo "Deploy ..."
    # deploy 自己写的python部署脚本
    # helm.yaml helm values配置文件
    # product-line 产品线
    # project-name 项目名称，如果多分支可以使用 project-name-$CI_COMMIT_REF_NAME
    # $(date "+%Y%m%d%H%M%S")-${CI_COMMIT_SHA:0:6} docker tag
    # Dockerfile 构建项目dockerfile
    - deploy -f helm.yaml product-line project-name $(date "+%Y%m%d%H%M%S")-${CI_COMMIT_SHA:0:6} Dockerfile 
  when: manual

deploy 部署脚本

  • 下面是 deploy 部分代码，获取脚本全部代码，请关注我的 YP小站 公众号并回复 获取deploy代码
#!/usr/bin/env python3

import sys
import os
import getopt
import yaml
import pykube

def usage():
    print("Usage: %s [ -b | --build-arg | -f | --file ] product-line project-name docker_images_version dockerfile_name --no-cache" % sys.argv[0], "\n", "or",
    "Usage: %s [ -b | --build-arg | -f | --file ] product-line project-name docker_images_version dockerfile_name" % sys.argv[0], "\n\n",
    "-b, --build-arg 声明Dockerfile中环境变量，参数可以指定多次，示例 A=b, ", "\n",
    "-f, --file 指定Helm values.yaml文件，参数只能指定一次，指定多个也只会取第一个值")

def check_item_exists(project, server_name, docker_images_version, *args):
    api = pykube.HTTPClient(pykube.KubeConfig.from_file("k8s config"))
    deploy = pykube.Deployment.objects(api).filter(namespace=project)
    os.environ['project'] = str(project)
    os.environ['server_name'] = str(server_name)
    os.environ['docker_images_version'] = str(docker_images_version)
    os.environ['s_name'] = str(server_name)
    if args:
        os.environ['helm_values_file'] = str(args[0])

    if os.system('helm status $server_name 1> /dev/null 2>&1') == 0:
        os.system("echo '\033[1;32;40m' 'Helm滚动发布、超时为5分钟、失败在本次基础上自动回滚上一个版本。' '\033[0m \n'")
        os.system('helm upgrade --timeout 300 --atomic --install $server_name --namespace $project \
        --set serverName=$server_name --set image.project=$project \
        --set serverFeatureNameReplace=$server_name \
        --set image.tag=$docker_images_version "Helm warehouse address" $helm_values_file')
        return
    else:
        # 兼容过去没有使用Helm部署
        for deployment in deploy:
            if server_name == str(deployment):
                # deployment进行滚动升级
                os.system('kubectl set image deployment/$server_name \
                 $server_name=harbor.example.com/$project/$server_name:$docker_images_version \
                 --namespace=$project')
                return

    # 部署新项目
    # 使用helm模板部署服务
    os.system('helm install --name $server_name --namespace $project \
    --set serverName=$server_name --set image.project=$project \
    --set serverFeatureNameReplace=$server_name \
    --set image.tag=$docker_images_version "Helm warehouse address" $helm_values_file')
    return

参考文献

https://blog.csdn.net/ygqygq2/article/details/85097857