README.md

The source code from this was adapted from the well-known Spring Music sample application.

The application has been refactored to use Spring Cloud Kubernetes as well as adopting the Event Sourcing and Outbox patterns.

Outbox

All events are captured in an Outbox table called outbox_events. There is a Debezium connector config file which will emit events into the outbox.${aggregateType}.events Kafka topic, where ${aggregateType} is driven by the domain event itself. Currently in this application there is only a single aggregate type, namely Album.

The outbox table is defined in the OutboxEvent JPA class. The resulting table should look like:

+-----------------+--------------+------+-----+---------+----------------+
| Field           | Type         | Null | Key | Default | Extra          |
+-----------------+--------------+------+-----+---------+----------------+
| event_id        | bigint       | NO   | PRI | NULL    | auto_increment |
| aggregate_id    | varchar(255) | NO   |     | NULL    |                |
| aggregate_type  | varchar(255) | NO   |     | NULL    |                |
| event_timestamp | datetime(6)  | NO   |     | NULL    |                |
| event_type      | varchar(255) | NO   |     | NULL    |                |
| payload         | json         | YES  |     | NULL    |                |
+-----------------+--------------+------+-----+---------+----------------+

Events will look something like

{
  "schema": {
    "type": "struct",
    "fields": [
      {
        "type": "string",
        "optional": true,
        "name": "io.debezium.data.Json",
        "version": 1,
        "field": "payload"
      },
      {
        "type": "string",
        "optional": true,
        "field": "eventType"
      },
      {
        "type": "int64",
        "optional": false,
        "field": "eventId"
      },
      {
        "type": "int64",
        "optional": true,
        "name": "io.debezium.time.MicroTimestamp",
        "version": 1,
        "field": "eventTimestamp"
      },
      {
        "type": "string",
        "optional": true,
        "field": "aggregateId"
      },
      {
        "type": "string",
        "optional": true,
        "field": "aggregateType"
      }
    ],
    "optional": false
  },
  "payload": {
    "payload": "{\"album\":{\"id\":\"5cc93f56-bc12-4e53-aeab-74c17ba027fa\",\"genre\":\"Blues\",\"title\":\"Couldn't Stand The Weather\",\"artist\":\"Stevie Ray Vaughan\",\"albumId\":null,\"trackCount\":0,\"releaseYear\":\"1984\"},\"eventType\":\"ALBUM_CREATED\"}",
    "eventType": "ALBUM_CREATED",
    "eventId": 29,
    "eventTimestamp": 1583344123573000,
    "aggregateId": "5cc93f56-bc12-4e53-aeab-74c17ba027fa",
    "aggregateType": "Album"
  }
}

{
  "schema": {
    "type": "struct",
    "fields": [
      {
        "type": "string",
        "optional": true,
        "name": "io.debezium.data.Json",
        "version": 1,
        "field": "payload"
      },
      {
        "type": "string",
        "optional": true,
        "field": "eventType"
      },
      {
        "type": "int64",
        "optional": false,
        "field": "eventId"
      },
      {
        "type": "int64",
        "optional": true,
        "name": "io.debezium.time.MicroTimestamp",
        "version": 1,
        "field": "eventTimestamp"
      },
      {
        "type": "string",
        "optional": true,
        "field": "aggregateId"
      },
      {
        "type": "string",
        "optional": true,
        "field": "aggregateType"
      }
    ],
    "optional": false
  },
  "payload": {
    "payload": "{\"eventType\":\"ALBUM_DELETED\",\"deletedAlbum\":{\"id\":\"0d2fc154-5734-4ee4-9c37-231f2c95877b\",\"genre\":\"Blues\",\"title\":\"Texas Flood\",\"artist\":\"Stevie Ray Vaughan\",\"albumId\":null,\"trackCount\":0,\"releaseYear\":\"1983\"}}",
    "eventType": "ALBUM_DELETED",
    "eventId": 30,
    "eventTimestamp": 1583344138851000,
    "aggregateId": "0d2fc154-5734-4ee4-9c37-231f2c95877b",
    "aggregateType": "Album"
  }
}

{
  "schema": {
    "type": "struct",
    "fields": [
      {
        "type": "string",
        "optional": true,
        "name": "io.debezium.data.Json",
        "version": 1,
        "field": "payload"
      },
      {
        "type": "string",
        "optional": true,
        "field": "eventType"
      },
      {
        "type": "int64",
        "optional": false,
        "field": "eventId"
      },
      {
        "type": "int64",
        "optional": true,
        "name": "io.debezium.time.MicroTimestamp",
        "version": 1,
        "field": "eventTimestamp"
      },
      {
        "type": "string",
        "optional": true,
        "field": "aggregateId"
      },
      {
        "type": "string",
        "optional": true,
        "field": "aggregateType"
      }
    ],
    "optional": false
  },
  "payload": {
    "payload": "{\"after\":{\"id\":\"0e44c7d9-f861-4823-a95e-64fe6bd23727\",\"genre\":\"Rock\",\"title\":\"Hotel California\",\"artist\":\"The Eagles\",\"albumId\":null,\"trackCount\":0,\"releaseYear\":\"1977\"},\"before\":{\"id\":\"0e44c7d9-f861-4823-a95e-64fe6bd23727\",\"genre\":\"Rock\",\"title\":\"Hotel California\",\"artist\":\"The Eagles\",\"albumId\":null,\"trackCount\":0,\"releaseYear\":\"1976\"},\"eventType\":\"ALBUM_UPDATED\"}",
    "eventType": "ALBUM_UPDATED",
    "eventId": 31,
    "eventTimestamp": 1583344145139000,
    "aggregateId": "0e44c7d9-f861-4823-a95e-64fe6bd23727",
    "aggregateType": "Album"
  }
}

Next will be working on wiring this together with a schema registry.

If you want to provision this app with all the Kafka clusters/Debezium setup on an OpenShift cluster, you can run the Ansible Playbook.