Home > Deploy > Change data capture (CDC) >

Change data capture (CDC) to Kafka Beta

Follow the steps below to connect a YugabyteDB cluster to use the Change Data Capture (CDC) API to send data changes to Apache Kafka. To learn about the change data capture (CDC) architecture, see Change data capture (CDC).

Prerequisites

YugabyteDB

Create a YugabyteDB cluster using the steps outlined in Manual Deployment.

Java

A JRE (or JDK), for Java 8 or 11, is installed.

Note

The Confluent Platform currently only supports Java 8 and 11. If you do not use one of these, an error message is generated and it will not start. For details related to the Confluent Platform, see Java version requirements.

Apache Kafka

A local install of the Confluent Platform should be up and running. The Confluent Platform includes Apache Kafka and additional tools and services (including Zookeeper and Avro), making it easy for you to quickly get started using the Kafka event streaming platform.

To get a local Confluent Platform (with Apache Kafka) up and running quickly, follow the steps in the Confluent Platform Quick Start (Local).

1. Create the source table

With your local YugabyteDB cluster running, create a table, called users, in the default database (yugabyte).

CREATE TABLE users (name text, pass text, id int, primary key (id));

2. Create Avro schemas

The Kafka Connect YugabyteDB Source Connector supports the use of Apache Avro schemas to serialize and deserialize tables. You can use the Schema Registry in the Confluent Platform to create and manage Avro schema files. For a step-by-step tutorial, see Schema Registry Tutorial.

Create two Avro schemas, one for the users table and one for the primary key of the table. After this step, you should have two files: table_schema_path.avsc and primary_key_schema_path.avsc.

You can use the following two Avro schema examples that will work with the users table you created.

table_schema_path.avsc:

{
  "type":"record",
  "name":"Table",
  "namespace":"org.yb.cdc.schema",
  "fields":[
  { "name":"name", "type":["null", "string"] },
  { "name":"pass", "type":["null", "string"] },
  { "name":"id", "type":["null", "int"] }
  ]
}

primary_key_schema_path.avsc:

{
  "type":"record",
  "name":"PrimaryKey",
  "namespace":"org.yb.cdc.schema",
  "fields":[
  { "name":"id", "type":["null", "int"] }
  ]
}

3. Start the Apache Kafka services

Create a Kafka topic.

./bin/kafka-topics --create --partitions 1 --topic users_topic --bootstrap-server localhost:9092 --replication-factor 1

Start the Kafka consumer service.

bin/kafka-avro-console-consumer --bootstrap-server localhost:9092 --topic users_topic --key-deserializer=io.confluent.kafka.serializers.KafkaAvroDeserializer     --value-deserializer=io.confluent.kafka.serializers.KafkaAvroDeserializer

4. Download the Kafka Connect YugabyteDB Source Connector

Download the Kafka Connect YugabyteDB Source Connector JAR file (yb-cdc-connector.jar).

$ wget -O yb-cdc-connector.jar https://github.com/yugabyte/yb-kafka-connector/blob/master/yb-cdc/yb-cdc-connector.jar?raw=true

5. Log to Kafka

Run the following command to start logging an output stream of data changes from the YugabyteDB cdc table to Apache Kafka.

java -jar yb-cdc-connector.jar \
--table_name yugabyte.cdc \
--topic_name cdc-test \
--table_schema_path table_schema_path.avsc \
--primary_key_schema_path primary_key_schema_path.avsc

The example above uses the following parameters:

--table_name — Specifies the namespace and table, where namespace is the database (YSQL) or keyspace (YCQL).
--master_addrs — Specifies the IP addresses for all of the YB-Master servers that are producing or consuming. Default value is 127.0.0.1:7100. If you are using a 3-node local cluster, then you need to specify a comma-delimited list of the addresses for all of your YB-Master servers.
topic_name — Specifies the Apache Kafka topic name.
table_schema_path — Specifies the location of the Avro file (.avsc) for the table schema.
primary_key_schema_path — Specifies the location of the Avro file (.avsc) for the primary key schema.

6. Write values and observe

In another window, write values to the table and observe the values on your Kafka output stream.