I’m open sourcing one of the modules I wrote for my Real-time Data Engineering class. We use Apache Spark and Apache Kafka to process data. Then, we show the data in real-time on a webpage using this JavaScript module to pull in data from Kafka via the Kafka REST interface.

The KafkaRESTHelper makes it easier to interface with the Kafka REST server and JQuery.

You can view the module at work in this video:

Configuration

This helper function makes a few assumptions about data and URLs. It assumes that the data is in Kafka and accessible by the Kafka REST server. It assumes the value of the payload is a valid JSON string.

Usage

Initialize the KafkaRESTHelper helper. The parameter is the URL of where to access the Kafka REST server. This example shows how to use Kafka REST behind a proxy:

var kafkaRESTHelper = new KafkaRESTHelper("/kafkarest/consumers/");

Next, you’ll create a createConsumerInstance by passing in the consumer group name, the topic name to consume, the function to call when data is retrieved, and the interval in millisecond to poll for data on the topic. This example shows an example call:

kafkaRESTHelper.createConsumerInstance("totalsbygameid", "totalsbygameid",
    byGameId, 10000)

The callback function will be called with a JavaScript object containing all of the data for the topic merged together. Here is an example function:

function byGameId(data) {
  console.log(data)
}

This function would power the rest of page. It might be used for charting or another function.

The module supports multiple calls to the createConsumerInstance function with different topics or the same topic.

TODOs

This code doesn’t delete the consumer instances automatically. I’d need to maintain a list of consumer nstances created and then register an onclose event to delete them.

Share This