Example – component anatomy

The following example shows how the component in the diagram at the beginning of this pattern could be configured and deployed using function-as-a-service, following the patterns discusses throughout this chapter. There is a function for publishing events following the Database-First Event Sourcing variant, a function for consuming events, and two RESTful functions for implementing a command and a query. Each function is its own independently scalable deployment unit, yet they are conveniently managed together as a whole. We will recognize this pattern as we dive into the higher-order patterns in Chapter 4, Boundary Patterns and Chapter 5, Control Patterns.

functions:
   command:
     handler: handler.command
     events:
       - http:
           path: items
           method: post
       - http:
           path: items/{id}
           method: put
     environment:
       TABLE_NAME:
         Ref: Table
   query:
     handler: handler.query
     events:
       - http:
           path: items/{id}
           method: get
     environment:
       TABLE_NAME:
         Ref: Table
   publish:
     handler: handler.publish
     events:
       - stream:
           type: dynamodb
           arn:
             Fn::GetAtt:
               - Table
               - StreamArn
           batchSize: 100
           startingPosition: TRIM_HORIZON
     environment:
       STREAM_NAME:
         Ref: Stream
   subscribe:
     handler: handler.consume
     events:
       - stream:
           type: kinesis
           arn:
             Fn::GetAtt:
               - Stream
               - Arn
           batchSize: 100
           startingPosition: TRIM_HORIZON
     environment:
       STREAM_NAME:
         Ref: Stream
       TABLE_NAME:
         Ref: Table