You are viewing documentation for an older version of Confluent Platform. For the latest, click here.

Replicator Quick Start

This guide will walk you through starting two Apache Kafka clusters and then starting a Replicator process to replicate data between them. Note that for tutorial purposes, we are runnining both clusters on the same machine. In order to do that, we jump through a hoop or two to make sure each cluster has its own ports and data directories. You will not need to perform these changes on the Zookeeper and Broker configuration if you are running in a normal environment where each cluster has its own servers.


Replicator Quick Start Configuration

Start the destination cluster

First, startup a ZooKeeper server. In this guide, we are assuming services will run on localhost.

# Start ZooKeeper. Run this command in its own terminal.
$ ./bin/zookeeper-server-start ./etc/kafka/

Next, startup a Kafka broker that will serve as our single node Kafka cluster.

# Start Kafka. Run this command in its own terminal.
$ ./bin/kafka-server-start ./etc/kafka/

For complete details on getting these services up and running see the quick start instructions for Confluent Platform.

Start the origin cluster

While we configured the destination cluster to run on default ports, we will need to run the origin cluster on a different port to avoid collisions. The Kafka in the origin cluster is configured on port 9082, Zookeeper is configured on 2171. Copy the configuration files to a temporary location and modify them so they do not conflict with the destination cluster.

#Copy the config files to /tmp
cp ./etc/kafka/ /tmp/
cp ./etc/kafka/ /tmp/

#Update the port numbers
sed -i '' -e "s/2181/2171/g" /tmp/
sed -i '' -e "s/9092/9082/g" /tmp/
sed -i '' -e "s/2181/2171/g" /tmp/
sed -i '' -e "s/#listen/listen/g" /tmp/

#Update data directories
sed -i '' -e "s/zookeeper/zookeeper_origin/g" /tmp/
sed -i '' -e "s/kafka-logs/kafka-logs-origin/g" /tmp/

From here, you can start up the origin cluster.

# Start ZooKeeper. Run this command in its own terminal.
$ ./bin/zookeeper-server-start /tmp/

# Start Kafka. Run this command in its own terminal.
$ ./bin/kafka-server-start /tmp/

Create a topic

Now, lets create a topic named “test-topic” in the origin cluster with the following command:

$ ./bin/kafka-topics --create --topic test-topic --replication-factor 1 --partitions 1 --zookeeper localhost:2171

Once we configure and run Replicator, this topic will get replicated to the destination cluster with the exact configuration we defined above. Note that for the sake of this example, we created a topic with just one partition. Replicator will work with any number of topics and partitions.

Configure and run Replicator

Confluent Replicator runs as a Connector in the Kafka Connect framework. In the quick start guide we will start a stand-alone Connect Worker process that runs Replicator as a Connector. For complete details on Connect see Kafka Connect.

The script that runs the stand-alone Connect Worker takes two configuration files. The first is the configuration for the Connect Worker itself and the second is the configuration for the Replicator.


Replicator is responsible for reading events from the origin cluster. It then passes the events to the Connect Worker which is responsible for writing the events to the destination cluster. Therefore we configure Replicator with information about the origin and the Worker with information about the destination.

We’ll start by configuring the the Connect Worker, and then configure the Replicator.

The Worker configuration file is ./etc/kafka/, edit the file and make sure it contains the addresses of brokers from the destination cluster. The default broker list will match the destination cluster we started earlier:

# Connect standalone worker configuration

Next, we will look at the Replicator configuration file, ./etc/kafka-connect-replicator/






Few of the configuration parameters here are important to understand and we’ll explain them here. You can read an explanation of all the configuration parameters in here.

  • key.converter and value.converter - Classes used to convert Kafka records to Connect’s internal format. The Connect Worker configuration specifies global converters and the default is JsonConverter. For Replication, however, no conversion is necessary. We just want to read bytes out of the origin cluster and write them to the destination with no changes. So we override the global converters with the ByteArrayConverter which just leaves the records as is.
  • src.kafka.bootstrap.servers - A list of brokers from the origin cluster
  • src.zookeeper.connect and dest.zookeeper.connect - Connection strings for Zookeeper in the origin and destination clusters respectively. These are used to replicate topic configuration from origin to destination.
  • topic.whitelist - An explicit list of the topics you want replicated. In this quick start, we will replicate the topic named “test-topic.”
  • topic.rename.format - A substitution string that is used to rename topics in the destination cluster. In the snippet above, we have used ${topic}.replica, where ${topic} will be substituted with the topic name from the origin cluster. That means that the test-topic we’re replicating from the origin cluster will be renamed to test-topic.replica in the destination cluster.

Once you update the quick start configuration, you can run the connector in a standalone Kafka Connect worker:

$ ./bin/connect-standalone ./etc/kafka/ ./etc/kafka-connect-replicator/

When the connector has finished initialization, it will check the origin cluster for topics that need to be replicated. In this case, it will find test-topic and will try to create the corresponding topic in the destination cluster. You can check this with the following command:

$ ./bin/kafka-topics --describe --topic test-topic.replica --zookeeper localhost:2181

Note that we’re checking the existence of test-topic.replica since test-topic was renamed according to our configuration. After verifying the topic’s existence, you should confirm that four partitions were created. In general, Replicator will ensure that the destination topic has as many partitions as the origin topic.

At any time after you’ve created the topic in the origin cluster, you can begin sending data to it using a Kafka producer to write to test-topic in the origin cluster. You can then confirm that the data has been replicated by consuming from test-topic.replica in the destination cluster. For example, to send a sequence of numbers using Kafka’s console producer, you can use the following command:

$ seq 10000 | ./bin/kafka-console-producer --topic test-topic --broker-list localhost:9082

You can then confirm delivery in the destination cluster using the console consumer:

$ ./bin/kafka-console-consumer --new-consumer --from-beginning --topic test-topic.replica --bootstrap-server localhost:9092

If you can see the numbers 1 to 10,000 appear in the consumer output - you can declare success. You now have working multi-cluster replication.