Edit

Development Environment

When developing KafkaJS, we run a Kafka cluster in a similar way to what is described in Running Kafka in Development, using docker and docker-compose. Before you proceed, make sure that you have both docker and docker-compose available.

KafkaJS is assuming that yarn is available globally, so if you haven't installed it yet: npm install --global yarn.

Running Kafka

For testing KafkaJS we use a multi-broker Kafka cluster as well as Zookeeper for authentication. To start the cluster and generate credentials, run the following from the root of the repository:

# This will run a Kafka cluster configured with your current IP
./scripts/dockerComposeUp.sh
./scripts/createScramCredentials.sh

This boots the Kafka cluster using the default docker-compose.yml file described in scripts/dockerComposeUp.sh. If you want to run a different version of Kafka, specify a different compose file using the COMPOSE_FILE environment variable:

COMPOSE_FILE="docker-compose.2_3.yml" ./scripts/dockerComposeUp.sh

If you run docker-compose -f docker-compose.2_3.yml ps (specify whichever compose file you used in the step above), you should see something like:

$ docker-compose -f docker-compose.2_3.yml ps
WARNING: The HOST_IP variable is not set. Defaulting to a blank string.
      Name                    Command               State                                   Ports
----------------------------------------------------------------------------------------------------------------------------------
kafkajs_kafka1_1   start-kafka.sh                   Up      0.0.0.0:9092->9092/tcp, 0.0.0.0:9093->9093/tcp, 0.0.0.0:9094->9094/tcp
kafkajs_kafka2_1   start-kafka.sh                   Up      0.0.0.0:9095->9095/tcp, 0.0.0.0:9096->9096/tcp, 0.0.0.0:9097->9097/tcp
kafkajs_kafka3_1   start-kafka.sh                   Up      0.0.0.0:9098->9098/tcp, 0.0.0.0:9099->9099/tcp, 0.0.0.0:9100->9100/tcp
kafkajs_zk_1       /bin/sh -c /usr/sbin/sshd  ...   Up      0.0.0.0:2181->2181/tcp, 22/tcp, 2888/tcp, 3888/tcp

The user credentials are listed in scripts/createScramCredentials.sh. You can also pass in the -h flag to this script for more details and controls.

You should now be able to connect to your cluster as such:

const fs = require('fs')
const ip = require('ip')

const { Kafka, CompressionTypes, logLevel } = require('./index')

const host = process.env.HOST_IP || ip.address()

const kafka = new Kafka({
  logLevel: logLevel.DEBUG,
  brokers: [`${host}:9094`, `${host}:9097`, `${host}:9100`],
  clientId: 'example-producer',
  ssl: {
    servername: 'localhost',
    rejectUnauthorized: false,
    ca: [fs.readFileSync('./testHelpers/certs/cert-signed', 'utf-8')],
  },
  sasl: {
    mechanism: 'plain',
    username: 'test',
    password: 'testtest',
  },
})

Visual Studio Code Integration

In order to better integrate with Visual Studio Code's Javascript Language Service, you can add a jsconfig.json file to the root of the project.

{
  "compilerOptions": {
    "baseUrl": ".",
    "module": "commonjs",
    "target": "es6",
    "paths": {
      "testHelpers": ["./testHelpers"]
    }
  },
  "include": [
    "src",
    "testHelpers"
  ]
}

This can help Visual Studio Code show type-hints even for modules that don't directly import each other: