Amazon Managed Service for Apache Flink was previously known as Amazon Kinesis Data Analytics for Apache Flink.

Use runtime properties in Managed Service for Apache Flink

You can use runtime properties to configure your application without recompiling your application code.

Manage runtime properties using the console

You can add, update, or remove runtime properties from your Managed Service for Apache Flink application using the AWS Management Console.

Manage runtime properties using the CLI

You can add, update, or remove runtime properties using the AWS CLI.

This section includes example requests for API actions for configuring runtime properties for an application. For information about how to use a JSON file for input for an API action, see Managed Service for Apache Flink API example code.

Add runtime properties when creating an application

The following example request for the CreateApplication action adds two runtime property groups (ProducerConfigProperties and ConsumerConfigProperties) when you create an application:

{
    "ApplicationName": "MyApplication",
    "ApplicationDescription": "my java test app",
    "RuntimeEnvironment": "FLINK-1_19",
    "ServiceExecutionRole": "arn:aws:iam::012345678901:role/MF-stream-rw-role",
    "ApplicationConfiguration": {
        "ApplicationCodeConfiguration": {
            "CodeContent": {
                "S3ContentLocation": {
                    "BucketARN": "arn:aws:s3:::ka-app-code-username",
                    "FileKey": "java-getting-started-1.0.jar"
                }
            },
            "CodeContentType": "ZIPFILE"
        },
        "EnvironmentProperties":  { 
         "PropertyGroups": [ 
            { 
               "PropertyGroupId": "ProducerConfigProperties",
               "PropertyMap" : {
                    "flink.stream.initpos" : "LATEST",
                    "aws.region" : "us-west-2",
                    "AggregationEnabled" : "false"
               }
            },
            { 
               "PropertyGroupId": "ConsumerConfigProperties",
               "PropertyMap" : {
                    "aws.region" : "us-west-2"
               }
            }
         ]
      }
    }
}

Add and update runtime properties in an existing application

The following example request for the UpdateApplication action adds or updates runtime properties for an existing application:

{
  "ApplicationName": "MyApplication",
  "CurrentApplicationVersionId": 2,
  "ApplicationConfigurationUpdate": {
    "EnvironmentPropertyUpdates": {
      "PropertyGroups": [ 
        { 
          "PropertyGroupId": "ProducerConfigProperties",
          "PropertyMap" : {
            "flink.stream.initpos" : "LATEST",
            "aws.region" : "us-west-2",
            "AggregationEnabled" : "false"
          }
        },
        { 
          "PropertyGroupId": "ConsumerConfigProperties",
          "PropertyMap" : {
            "aws.region" : "us-west-2"
          }
        }
      ]
    }
  }
}

Remove runtime properties

The following example request for the UpdateApplication action removes all runtime properties and property groups from an existing application:

{
  "ApplicationName": "MyApplication",
  "CurrentApplicationVersionId": 3,
  "ApplicationConfigurationUpdate": {
    "EnvironmentPropertyUpdates": {
      "PropertyGroups": []
    }
  }
}

Access runtime properties in a Managed Service for Apache Flink application

You retrieve runtime properties in your Java application code using the static KinesisAnalyticsRuntime.getApplicationProperties() method, which returns a Map<String, Properties> object.

The following Java code example retrieves runtime properties for your application:

 Map<String, Properties> applicationProperties = KinesisAnalyticsRuntime.getApplicationProperties();

You retrieve a property group (as a Java.Util.Properties object) as follows:

Properties consumerProperties = applicationProperties.get("ConsumerConfigProperties");

You typically configure an Apache Flink source or sink by passing in the Properties object without needing to retrieve the individual properties. The following code example demonstrates how to create an Flink source by passing in a Properties object retrieved from runtime properties:

private static FlinkKinesisProducer<String> createSinkFromApplicationProperties() throws IOException {
  Map<String, Properties> applicationProperties = KinesisAnalyticsRuntime.getApplicationProperties();
  FlinkKinesisProducer<String> sink = new FlinkKinesisProducer<String>(new SimpleStringSchema(),
    applicationProperties.get("ProducerConfigProperties"));

  sink.setDefaultStream(outputStreamName);
  sink.setDefaultPartition("0");
  return sink;
}

For code examples, see Examples for creating and working with Managed Service for Apache Flink applications.