通过云形成、代码构建和代码管道将Python包部署到AWS Lambda

Question

我想为我的AWS基础设施和AWS Lambda函数设置一个CI/CD管道。我们的想法是让一切都在代码中，版本控制和自动化。我只想git将推送到一个存储库，让CodePipeline从那里接管，更新我的基础设施，运行测试，如果成功，用最新的代码更新Lambda函数。

我的CloudFormation模板基于这个优秀的示例。看起来是这样的：

AWSTemplateFormatVersion: 2010-09-09
Description: playground pipeline 1
Parameters:
  SourceRepositoryName:
    Type: String
    Default: lambda-playground
  SourceBranchName:
    Type: String
    Default: master

Resources:
  ArtifactsBucket:
    Type: AWS::S3::Bucket
    DependsOn: CloudFormationRole
    DeletionPolicy: Delete
    Properties:
      BucketName: lambda-playground-artifacts

  CodeBuildRole:
    Type: AWS::IAM::Role
    DependsOn: CloudFormationRole
    Properties:
      AssumeRolePolicyDocument:
        Version: 2012-10-17
        Statement:
          - Effect: Allow
            Action:
              - sts:AssumeRole
            Principal:
              Service:
                - codebuild.amazonaws.com
      Policies:
        - PolicyName: ServiceRole
          PolicyDocument:
            Version: 2012-10-17
            Statement:
              - Sid: CloudWatchWriteLogsPolicy
                Effect: Allow
                Action:
                  - logs:CreateLogGroup
                  - logs:CreateLogStream
                  - logs:PutLogEvents
                Resource: '*'
              - Sid: CodeCommitPullPolicy
                Effect: Allow
                Action:
                  - codecommit:GitPull
                Resource: '*'
              - Sid: S3GetObjectPolicy
                Effect: Allow
                Action:
                  - s3:GetObject
                  - s3:GetObjectVersion
                Resource: '*'
              - Sid: S3PutObjectPolicy
                Effect: Allow
                Action:
                  - s3:PutObject
                Resource: '*'

  CodePipelineRole:
    Type: AWS::IAM::Role
    DependsOn: CloudFormationRole
    Properties:
      AssumeRolePolicyDocument:
        Version: 2012-10-17
        Statement:
          - Effect: Allow
            Action:
              - sts:AssumeRole
            Principal:
              Service:
                - codepipeline.amazonaws.com
      ManagedPolicyArns:
        - arn:aws:iam::aws:policy/AdministratorAccess

  CloudFormationRole:
    Type: AWS::IAM::Role
    Properties:
      AssumeRolePolicyDocument:
        Version: 2012-10-17
        Statement:
          - Effect: Allow
            Action:
              - sts:AssumeRole
            Principal:
              Service:
                - cloudformation.amazonaws.com
      ManagedPolicyArns:
        - arn:aws:iam::aws:policy/AdministratorAccess

  CodeCommitRepository:
    Type: AWS::CodeCommit::Repository
    Properties:
      RepositoryName: !Ref SourceRepositoryName

  CodeBuildProject:
    Type: AWS::CodeBuild::Project
    DependsOn: CloudFormationRole
    Properties:
      Description: A playground of Lambda
      Artifacts:
        Type: CODEPIPELINE
      Environment:
        ComputeType: BUILD_GENERAL1_SMALL
        Image: aws/codebuild/python:2.7.12
        Type: LINUX_CONTAINER
      Name: lambda-playground
      ServiceRole: !GetAtt CodeBuildRole.Arn
      Source:
        Type: CODEPIPELINE
      TimeoutInMinutes: 5

  CodePipeline:
    Type: AWS::CodePipeline::Pipeline
    Properties:
      ArtifactStore:
        Type: S3
        Location: !Ref ArtifactsBucket
      Name: !Ref AWS::StackName
      RestartExecutionOnUpdate: true
      RoleArn: !GetAtt CodePipelineRole.Arn
      Stages:
        - Name: Source
          Actions:
            - Name: Source
              ActionTypeId:
                Category: Source
                Owner: AWS
                Provider: CodeCommit
                Version: 1
              Configuration:
                RepositoryName: !Ref SourceRepositoryName
                BranchName: !Ref SourceBranchName
              OutputArtifacts:
                - Name: SourceOutput
        - Name: PipelineDeploy
          Actions:
            - Name: UpdatePipeline
              ActionTypeId:
                Category: Deploy
                Owner: AWS
                Provider: CloudFormation
                Version: 1
              Configuration:
                ActionMode: CREATE_UPDATE
                Capabilities: CAPABILITY_IAM
                RoleArn: !GetAtt CloudFormationRole.Arn
                StackName: !Ref AWS::StackName
                TemplatePath: SourceOutput::infra.yml
              InputArtifacts:
                - Name: SourceOutput
        - Name: Build
          Actions:
            - Name: BuildAndTest
              ActionTypeId:
                Category: Build
                Owner: AWS
                Provider: CodeBuild
                Version: 1
              Configuration:
                ProjectName: !Ref CodeBuildProject
              InputArtifacts:
                - Name: SourceOutput
              OutputArtifacts:
                - Name: BuildOutput

  LambdaFunction:
    Type: AWS::Lambda::Function
    Properties:
      Code:
        S3Bucket: !Ref ArtifactsBucket
        S3Key: !Ref BuildOutput # DOES NOT WORK
      FunctionName: playground-fc
      Handler: src.main.handler
      # TODO: Role: foo
      Runtime: python2.7

Outputs:
  ArtifactsBucketURL:
    Description: Artifacts bucket URL
    Value: !GetAtt ArtifactsBucket.WebsiteURL
  RepositoryURL:
    Description: SSH URL of the repository
    Value: !GetAtt CodeCommitRepository.CloneUrlSsh

所以我有一个CodePipeline，有3个阶段-Source，它从CodeCommit回购中获取代码，PipelineDeploy，它在必要时更新我的云形成堆栈和Build，它运行配置的CodeBuild项目。

我的建筑规范。yml在此：

version: 0.1

phases:
  install:
    commands:
      - pip install -r requirements.txt -t lib
  pre_build:
    commands:
      - python lib/pytest.py src
artifacts:
  type: zip
  files:
    - src/**/*
    - lib/**/*

它只是安装必要的库，通过pytest运行测试，并创建一个部署zip。然后，此zip文件是Build阶段的OutputArtifact，并存储在ArtifactsBucket中。但是，每次它都会得到一个唯一的名称（例如，dfVV6Uh），这很有意义，但我不知道如何在lambda函数中引用它-

所以我的问题是，如何创建一个堆栈/管道，在完成所有步骤后，将最新版本部署到AWS Lambda函数？有没有一种方法可以使用CodeDeploy来实现这一点？这里的最佳实践是什么？




Answer 1

有一个例子说明了如何实现类似的功能（通过CodePipeline/CodeBuild部署lambda函数）。http://docs.aws.amazon.com/lambda/latest/dg/automating-deployment.html

这个例子是针对NodeJS中编写的lambda函数，但是基本思想是一样的。在您通过CodeBuild构建工件并让CodePipeline在阶段内管理工件传播后，您可以使用CloudForm部署/更新lambda函数。

让我知道这是否有帮助。

Answer 2

您可以使用参数覆盖和GetArtifactAtt以及ObjectKey属性动态提供工件的名称。由AWS CodePipeline生成的压缩包，用于您的CloudFormation部署操作。

使用您的示例，您的updatePilineCloudFormation部署操作的配置如下所示：

Configuration:
  ActionMode: CREATE_UPDATE
  Capabilities: CAPABILITY_IAM
  RoleArn: !GetAtt CloudFormationRole.Arn
  StackName: !Ref AWS::StackName
  TemplatePath: SourceOutput::infra.yml
  ParameterOverrides:
    {
      "LambdaKey" : { "Fn::GetArtifactAtt" : ["LambdaFunctionSource", "ObjectKey"]}
    }
InputArtifacts:
- Name: SourceOutput
- Name: BuildOutput

然后，声明并引用CloudForm堆栈模板中的LambdaKeyParameter：

Parameters:
  LambdaKey:
    Type: String
  # ...
Resources:
  LambdaFunction:
    Type: AWS::Lambda::Function
    Properties:
      Code:
        S3Bucket: !Ref ArtifactsBucket
        S3Key: !Ref LambdaKey
      # ...