template.yaml

AWSTemplateFormatVersion: '2010-09-09'
Description: >
  Example PySpark AWS Glue Job w/ AWS SAM

Resources:
  MySourceBucket:
    Type: "AWS::S3::Bucket"
    Properties: 
      BucketName: "example-source-glue-pyspark-demo"
  
  MyTargetBucket:
    Type: "AWS::S3::Bucket"
    Properties: 
      BucketName: "example-target-glue-pyspark-demo"

  ExampleSAMGlueRole:
    Type: "AWS::IAM::Role"
    Properties:
      RoleName: "ExampleSAMGlueRoleName"
      AssumeRolePolicyDocument:
        Version: "2012-10-17"
        Statement:
          -
            Effect: "Allow"
            Principal:
              Service:
                - "glue.amazonaws.com"
            Action:
              - "sts:AssumeRole"
      Path: "/"
      Policies:
        -
          PolicyName: "ExampleSAMGluePolicy"
          PolicyDocument:
            Version: "2012-10-17"
            Statement:
              -
                Effect: "Allow"
                Action: "s3:*Object"
                Resource: 
                  - !Sub "arn:aws:s3:::${MySourceBucket}/*"
                  - !Sub "arn:aws:s3:::${MyTargetBucket}/*"
              - 
                Effect: "Allow"
                Action: "s3:ListBucket"
                Resource:
                  - !Sub  "arn:aws:s3:::${MySourceBucket}"
                  - !Sub  "arn:aws:s3:::${MyTargetBucket}"
              - 
                Effect: "Allow"
                Action: "*"
                Resource:
                  - "arn:aws:s3:::aws-sam-cli-managed-default*" # this is the default bucket where sam cli stores your code/templates


  ExampleSAMGlueJobETL:
    Type: "AWS::Glue::Job"
    Properties: 
      Command: 
          Name: "glueetl" #for Spark jobs name must be glueetl per docs
          ScriptLocation: ./glue-script/example-pyspark-glueetl.py #Reference to the file we are uploading
      Description: "Performs an ETL Job on Source Bucket w/ Time Stamps"
      GlueVersion: "3.0"
      MaxCapacity: 10
      MaxRetries: 0
      Name: "ExampleDemoGlueJobETL"
      Role: !Ref ExampleSAMGlueRole
      Timeout: 3