AzkarraApplication

The AzkarraApplication is the top-level concept that is used to bootstrap an Azkarra Streams application.

Its main responsibility is to initialize an AzkarraContext using a user-defined configuration which describes the StreamsExecutionEnvironments and the streams applications to be executed.

The AzkarraApplication is also responsible for deploying an embedded HTTP server (if enable)

1 Creating a new AzkarraApplication

AzkarraApplication application = new AzkarraApplication();

2 Auto-configuration application

The AzkarraApplication allows you to automatically configure and assemble the internal AzkarraContext, StreamsExecutionEnvironment and TopologyProvider instances using the AzkarraConf class.

AzkarraConf azkarraConf = AzkarraConf.create("application");
AzkarraApplication application = new AzkarraApplication()
        .setConfiguration(azkarraConf);

2.1 Configuring AzkarraContext

azkarra {
  // The context configuration
  context {
    // The default configuration for streams application.
    streams {
      bootstrap.servers = "localhost:9092"
      default.key.serde = "org.apache.kafka.common.serialization.Serdes$StringSerde"
      default.value.serde = "org.apache.kafka.common.serialization.Serdes$StringSerde"
    }
  }
  ...
}

2.2 Configuring components to register

Complete example (using HOCON format):

azkarra {
  ...	
  // Manually defines the providers to be registered
  components = [
    io.streamthoughts.azkarra.example.topology.WordCountTopology
  ]

}

2.3 Configuring StreamsExecutionEnvironment and TopologyProvider

The azkarra.environments property allows you to list the StreamsExecutionEnvironment to initialize and configure.

Here are the properties to use:

Moreover, you can declared the list of TopologyProvider to be registered into a StreamsExecutionEnvironment using the following properties:

azkarra {
  ...	
  // Create a default environment for running the WordCountTopology
  environments = [
    {
      name: "__default"
      config = {}
      jobs = [
        {
          name = "word-count-demo"
          description = "Kafka Streams WordCount Demo"
          topology =  "WordCountTopology"
          config = {}
        }
      ]
    }
  ]
}

3 Enabling component-scan

Azkarra provides a simple mechanism to search for components present in the classpath or in external directories.

Any class annotated with @Component or @Factory will be scanned and automatically registered to the AzkarraContext.

The component-scan can be enable programmatically through the method AzkarraApplication#setEnableComponentScan. To automatically start all registered TopologyProviver, you must also enable auto-start.

Code snippet for enabling component scan :

application.setEnableComponentScan(true);

4 Auto-starting topologies

By enabling auto-start all TopologyProvider that was registered during component-scanning will be automatically added to the given StreamExecutionEnvironment.

application.setAutoStart(true, "__default");

5 The @AzkarraStreamsApplication annotation

Another way to initialize and configure an AzkarraApplication is to use annotations. Therefore, you can simply annotated you main class with @AzkarraStreamsApplication and initialize your AzkarraApplication as follows:

@AzkarraStreamsApplication
public class SimpleStreamsApp {

    public static void main(final String[] args) {
        AzkarraApplication.run(SimpleStreamsApp.class, args);
    }
}

The @AzkarraStreamsApplication is a convenience annotation that adds all of the following:

• @EnableAutoConfig: Tells Azkarra to auto-configure internal AzkarraContext
• @EnableAutoStart: Tells Azkarra to automatically start any registered TopologyProvider using the default environment.
• @EnableEmbeddedHttpServer: Tells Azkarra to start the embedded http-server.
• @ComponentScan: Tells Azkarra to look for classes annotated with @Component in the current package.