当前位置：首页 > news >正文

aws(学习笔记第三十八课) codepipeline-build-deploy-github-manual

news 来源：原创 2025/6/27 11:48:15

文章目录

aws(学习笔记第三十八课) codepipeline-build-deploy-github-manual
学习内容：
- 1. 整体架构
- - 1.1 代码链接
  - 1.2 全体处理架构
- 2. 代码分析
- - 2.1 创建`ImageRepo`，并设定给`FargateTaskDef`
  - 2.2 创建`CodeBuild project`
  - 2.3 对`CodeBuild project`赋予权限（`ECR`的`image repo`）
  - 2.4 生成`lambda`函数触发`codebuild`的首次执行
  - - 2.4.1 生成`lambda`函数触发`codebuild`
    - 2.4.2 `lambda`函数触发`codebuild`的代码
    - 2.4.3 `AWS Custom Resource`的代码
  - 2.5 创建`vpc`和`ecs cluster`
  - - 2.5.1 创建`vpc`
    - 2.5.2 创建`蓝绿部署`
  - 2.5 创建`Application Loader Balancer`
  - 2.6 创建`fargate service`
  - 2.6.1 创建`fargate service`
  - 2.6.1 在`fargate service`上进行`application target group`的`attach`
  - 2.7 创建`code pipeline`
  - - 2.7.2 `codepipe`整体架构
    - 2.7.1 创建`pipeline artifacts`
    - 2.7.2 创建`source stage`
    - 2.7.3 创建`build stage`
    - 2.7.4 创建`deploy stage`
  - 2.8 创建`code pipeline`
  - - 2.8.1 为`code build`设定`github`的`confidential`
    - 2.8.2 最终创建`pipeline`
- 3 执行`CDK`

aws(学习笔记第三十八课) codepipeline-build-deploy-github-manual

使用codepipeline-build-deploy

学习内容：

使用codepipeline，★代码库不采用codecommit，而是github
使用codebuild
使用codedeploy

1. 整体架构

1.1 代码链接

代码链接(codepipeline-build-deploy-github-manual)
同时可以参照aws(学习笔记第二十课) codecommit以及codedeploy进行开发

1.2 全体处理架构

在这里插入图片描述

2. 代码分析

2.1 创建`ImageRepo`，并设定给`FargateTaskDef`

        # Creates an Elastic Container Registry (ECR) image repository
        image_repo = ecr.Repository(self, "ImageRepo")

        # Creates a Task Definition for the ECS Fargate service
        fargate_task_def = ecs.FargateTaskDefinition(self, "FargateTaskDef")
        fargate_task_def.add_container(
            "Container",
            container_name="web",
            image=ecs.ContainerImage.from_ecr_repository(image_repo),
            port_mappings=[{"containerPort": 80}]
        )

注意，这里并没有指定具体的image的URI以及tag，这里只是指定了Image Repo(ECR)。后面会在CodeBuildStage阶段build出来docker image，并且接在CodeDeployStage阶段将该docker image部署到Fargate Service。
在这里插入图片描述

2.2 创建`CodeBuild project`

# CodeBuild project that builds the Docker image
        build_image = codebuild.Project(
            self, "BuildImage",
            build_spec=codebuild.BuildSpec.from_source_filename(
                "buildspec.yaml"),
            source=codebuild.Source.git_hub(
                owner="bufanli",    # TODO: Replace with your GitHub username
                repo="web_app",   # TODO: Replace with your GitHub repository name
                branch_or_ref="main",
            ),
            environment=codebuild.BuildEnvironment(
                privileged=True
            ),
            environment_variables={
                "AWS_ACCOUNT_ID": codebuild.BuildEnvironmentVariable(value=os.getenv('CDK_DEFAULT_ACCOUNT') or ""),
                "REGION": codebuild.BuildEnvironmentVariable(value=os.getenv('CDK_DEFAULT_REGION') or ""),
                "IMAGE_TAG": codebuild.BuildEnvironmentVariable(value="latest"),
                "IMAGE_REPO_NAME": codebuild.BuildEnvironmentVariable(value=image_repo.repository_name),
                "REPOSITORY_URI": codebuild.BuildEnvironmentVariable(value=image_repo.repository_uri),
                "TASK_DEFINITION_ARN": codebuild.BuildEnvironmentVariable(value=fargate_task_def.task_definition_arn),
                "TASK_ROLE_ARN": codebuild.BuildEnvironmentVariable(value=fargate_task_def.task_role.role_arn),
                "EXECUTION_ROLE_ARN": codebuild.BuildEnvironmentVariable(value=fargate_task_def.execution_role.role_arn)
            }

在这里插入图片描述
这里，设定build_spec文件，在app/buildspec.yml中，

设定pre_build登录ecr (‌Amazon Elastic Container Registry（ECR)，方便之后的向ecr进行镜像的登录。

    commands:
      - echo Logging in to Amazon ECR...
      - aws --version
      - aws ecr get-login-password --region $REGION | docker login --username AWS --password-stdin $AWS_ACCOUNT_ID.dkr.ecr.$REGION.amazonaws.com

进行build处理
这里的

    commands:
      - echo Building the Docker image...
      - docker build -t $IMAGE_REPO_NAME:$IMAGE_TAG .
      - docker tag $IMAGE_REPO_NAME:$IMAGE_TAG $AWS_ACCOUNT_ID.dkr.ecr.$REGION.amazonaws.com/$IMAGE_REPO_NAME:$IMAGE_TAG

注意，这里的环境给build_spec.yml传递变量都是通过env(环境变量)来传递的。
可以看到，这里build执行docker build命令，进行docker image的build。
在这里插入图片描述

进行post build处理
这里的变量也同样是在codebuild中，在cdk的codebuild.project中，通过变量传递过来的。

    commands:
      - echo Pushing the Docker image...
      - docker push $AWS_ACCOUNT_ID.dkr.ecr.$REGION.amazonaws.com/$IMAGE_REPO_NAME:$IMAGE_TAG
      - echo Container image to be used $REPOSITORY_URI:$IMAGE_TAG
      - sed -i "s|REPOSITORY_URI|${REPOSITORY_URI}|g" taskdef.json
      - sed -i "s|IMAGE_TAG|${IMAGE_TAG}|g" taskdef.json
      - sed -i "s|TASK_ROLE_ARN|${TASK_ROLE_ARN}|g" taskdef.json
      - sed -i "s|EXECUTION_ROLE_ARN|${EXECUTION_ROLE_ARN}|g" taskdef.json
      - sed -i "s|TASK_DEFINITION_ARN|${TASK_DEFINITION_ARN}|g" appspec.yaml
      - cat appspec.yaml && cat taskdef.json

2.3 对`CodeBuild project`赋予权限（`ECR`的`image repo`）

        # Grants CodeBuild project access to pull/push images from/to ECR repo
        image_repo.grant_pull_push(build_image)

2.4 生成`lambda`函数触发`codebuild`的首次执行

2.4.1 生成`lambda`函数触发`codebuild`

整体架构如下：

aws custom resource triggers lambda on create event
lambda executes code build project
code build project starts source stage,build stage and deploy stage

2.4.2 `lambda`函数触发`codebuild`的代码

CDK代码
注意，这里赋予了lamba权限，允许执行code build

        # Lambda function that triggers CodeBuild image build project
        trigger_code_build = lambda_.Function(
            self, "BuildLambda",
            architecture=lambda_.Architecture.ARM_64,
            code=lambda_.Code.from_asset("lambda"),
            handler="trigger-build.handler",
            runtime=lambda_.Runtime.NODEJS_18_X,
            environment={
                "CODEBUILD_PROJECT_NAME": build_image.project_name,
                "REGION": os.getenv('CDK_DEFAULT_REGION') or ""
            },
            # Allows this Lambda function to trigger the buildImage CodeBuild project
            initial_policy=[
                iam.PolicyStatement(
                    effect=iam.Effect.ALLOW,
                    actions=["codebuild:StartBuild"],
                    resources=[build_image.project_arn]
                )
            ]
        )

lambda函数代码

const {
  CodeBuildClient,
  StartBuildCommand,
} = require("@aws-sdk/client-codebuild");

exports.handler = async (event) => {
  const region = process.env.REGION;
  const buildProjectName = process.env.CODEBUILD_PROJECT_NAME;

  const codebuild = new CodeBuildClient({ region: region });
  const buildCommand = new StartBuildCommand({ projectName: buildProjectName });

  console.log("Triggering CodeBuild Project...");
  const buildResponse = await codebuild.send(buildCommand);
  console.log(buildResponse);

  return {
    statusCode: 200,
    body: "CodeBuild Project building...",
  };
};

2.4.3 `AWS Custom Resource`的代码

        # Triggers a Lambda function using AWS SDK
        trigger_lambda = custom.AwsCustomResource(
            self, "BuildLambdaTrigger",
            install_latest_aws_sdk=True,
            policy=custom.AwsCustomResourcePolicy.from_statements([
                iam.PolicyStatement(
                    effect=iam.Effect.ALLOW,
                    actions=["lambda:InvokeFunction"],
                    resources=[trigger_code_build.function_arn],
                )
            ]),
            on_create={
                "service": "Lambda",
                "action": "invoke",
                "physical_resource_id": custom.PhysicalResourceId.of("id"),
                "parameters": {
                    "FunctionName": trigger_code_build.function_name,
                    "InvocationType": "Event",
                },
            },
            on_update={
                "service": "Lambda",
                "action": "invoke",
                "parameters": {
                    "FunctionName": trigger_code_build.function_name,
                    "InvocationType": "Event",
                },
            }
        )

2.5 创建`vpc`和`ecs cluster`

2.5.1 创建`vpc`

        # Creates VPC for the ECS Cluster
        cluster_vpc = ec2.Vpc(
            self, "ClusterVpc",
            ip_addresses=ec2.IpAddresses.cidr(cidr_block="10.75.0.0/16")
        )

        # Deploys the cluster VPC after the initial image build triggers
        cluster_vpc.node.add_dependency(trigger_lambda)

注意，这里创建vpc需要依赖trigger _lambda，只有当trigger lambda的aws custom resource创建好了，才能创建vpc。

2.5.2 创建`蓝绿部署`

蓝绿部署的整体架构
- 【1】利用CodeDepoly服务，开始进行application deploy
- 【2】生成导入新的application的Green Group
- 【3】在Application对开发者设定对Green Group的访问的listener
- 【4】开发者对Green Group进行动作确认，确保生产环境上没有问题
- 【5】在Application对一般用户设定对访问的listener，让其指向Green Group，设定Blue Group为新的Green Group
- 【6】删除原来的Blue Group。这里取决于设定的策略，可能不是马上删除。

创建ApplicationTargetGroup(blue)蓝色目标组

        target_group_blue = elb.ApplicationTargetGroup(
            self, "BlueTargetGroup",
            target_group_name="alb-blue-tg",
            target_type=elb.TargetType.IP,
            port=80,
            vpc=cluster_vpc
        )

在这里插入图片描述

创建ApplicationTargetGroup(green)绿色目标组

    # Creates a new green Target Group
    target_group_green = elb.ApplicationTargetGroup(
        self, "GreenTargetGroup",
        target_group_name="alb-green-tg",
        target_type=elb.TargetType.IP,
        port=80,
        vpc=cluster_vpc
    )

在这里插入图片描述

2.5 创建`Application Loader Balancer`

注意，这里 default_target_groups=[target_group_blue]默认的target group设成了blue group。

       # Creates a new green Target Group
        target_group_green = elb.ApplicationTargetGroup(
            self, "GreenTargetGroup",
            target_group_name="alb-green-tg",
            target_type=elb.TargetType.IP,
            port=80,
            vpc=cluster_vpc
        )

        # Creates a Security Group for the Application Load Balancer (ALB)
        albSg = ec2.SecurityGroup(
            self, "AlbSecurityGroup",
            vpc=cluster_vpc,
            allow_all_outbound=True
        )
        albSg.add_ingress_rule(
            peer=ec2.Peer.any_ipv4(),
            connection=ec2.Port.tcp(80),
            description="Allows access on port 80/http",
            remote_rule=False
        )

        # Creates a public ALB
        public_alb = elb.ApplicationLoadBalancer(
            self, "PublicAlb",
            vpc=cluster_vpc,
            internet_facing=True,
            security_group=albSg
        )

        # Adds a listener on port 80 to the ALB
        alb_listener = public_alb.add_listener(
            "AlbListener80",
            open=False,
            port=80,
            default_target_groups=[target_group_blue]
        )

2.6 创建`fargate service`

2.6.1 创建`fargate service`

注意，这里直接在创建fargate service的时候，直接创建ecs cluster。并且将task def和ecs cluster联系起来。

        # Creates an ECS Fargate service
        fargate_service = ecs.FargateService(
            self, "FargateService",
            desired_count=1,
            service_name="fargate-frontend-service",
            task_definition=fargate_task_def,
            cluster=ecs.Cluster(
                self, "EcsCluster",
                enable_fargate_capacity_providers=True,
                vpc=cluster_vpc
            ),
            # Sets CodeDeploy as the deployment controller
            deployment_controller=ecs.DeploymentController(
                type=ecs.DeploymentControllerType.CODE_DEPLOY
            ),
        )

2.6.1 在`fargate service`上进行`application target group`的`attach`

 # Adds the ECS Fargate service to the ALB target group
fargate_service.attach_to_application_target_group(target_group_blue)

2.7 创建`code pipeline`

2.7.2 `codepipe`整体架构

在这里插入图片描述

2.7.1 创建`pipeline artifacts`

        # Creates new pipeline artifacts
        source_artifact = pipeline.Artifact("SourceArtifact")
        build_artifact = pipeline.Artifact("BuildArtifact")

在这里插入图片描述

source_artifact是由source_stage作成，由build_stage进行使用
build_artifact是由build_stage作成，由deploy_stage进行使用，并且包括：
- appspec.yml
- taskdef.json

2.7.2 创建`source stage`

        # Creates the source stage for CodePipeline
        source_stage = pipeline.StageProps(
            stage_name="Source",
            actions=[
                pipelineactions.GitHubSourceAction(
                    action_name="GitHub",
                    owner='XXXXXX',    # TODO: Replace with your GitHub username
                    repo='web_app',   # TODO: Replace with your GitHub repository name
                    branch="main",
                    oauth_token=SecretValue.secrets_manager("github-access-token-secret"),
                    output=source_artifact,
                )
            ]
        )

这里，需要设定几个地方：
- owner 这里是github工程的用户名，这里就是自己的github账号。
- repo这里是github库的名字。
- oauth_token，因为需要将aws的url(callback)，注册给github，所以需要 oauth_token进行验证。因为aws的source stage需要检测到github的commit，自动触发code pipeline，所以需要注册url(callback)。
  下面的就是讲本次example程序中的application的部分push到github，创建一个示例的web application。
github的webhooks和codepipeline交互
在cdk中练习使用aws secret manager

2.7.3 创建`build stage`

        # Creates the build stage for CodePipeline
        build_stage = pipeline.StageProps(
            stage_name="Build",
            actions=[
                pipelineactions.CodeBuildAction(
                    action_name="DockerBuildPush",
                    input=pipeline.Artifact("SourceArtifact"),
                    project=build_image,
                    outputs=[build_artifact]
                )
            ]
        )

这里使用source stage生成的source artifact，进行docker build，之后将docker image进行push到aws ecr。
project设定这里要设定成前面创建的build_image

2.7.4 创建`deploy stage`

整体构造

创建deployment group

# Creates a new CodeDeploy Deployment Group
        deployment_group = codedeploy.EcsDeploymentGroup(
            self, "CodeDeployGroup",
            service=fargate_service,
            # Configurations for CodeDeploy Blue/Green deployments
            blue_green_deployment_config=codedeploy.EcsBlueGreenDeploymentConfig(
                listener=alb_listener,
                blue_target_group=target_group_blue,
                green_target_group=target_group_green
            )
        )

创建deploy stage

        # Creates the deploy stage for CodePipeline
        deploy_stage = pipeline.StageProps(
            stage_name="Deploy",
            actions=[
                pipelineactions.CodeDeployEcsDeployAction(
                    action_name="EcsFargateDeploy",
                    app_spec_template_input=build_artifact,
                    task_definition_template_input=build_artifact,
                    deployment_group=deployment_group
                )
            ]
        )

2.8 创建`code pipeline`

2.8.1 为`code build`设定`github`的`confidential`

        # Store Github credentials to CodeBuild
        codebuild.GitHubSourceCredentials(self, "CodeBuildGitHubCreds",
            access_token=SecretValue.secrets_manager("github-access-token-secret")
        )

2.8.2 最终创建`pipeline`

        # Creates an AWS CodePipeline with source, build, and deploy stages
        pipeline.Pipeline(
            self, "BuildDeployPipeline",
            pipeline_name="ImageBuildDeployPipeline",
            stages=[source_stage, build_stage, deploy_stage]
        )