038ee4576a
- get rid of topic creator and scala deps - use boot properties - demonstrate using the DSL to dynamically add adapters
Apache Kafka Sample
This example demonstrates the use Kafka Outbound Channel Adapter and Message-Driven Channel Adapter.
It uses Java configuration for the adapters.
Running the sample
Start Apache Zookeeper and Apache Kafka according to the documentation for the Apache Kafka project.
$ gradlew :kafka:run
This will package the application and run it using the Gradle Application Plugin
Using an IDE such as SpringSource Tool Suite™ (STS)
In STS (Eclipse), go to package org.springframework.integration.samples.kafka, right-click Application and select Run as --> Java Application (or Spring Boot Application).
Output
The application sends 10 messages (foo0 ... foo9) to a kafka topic si.topic; this must exist or the broker must be configured to auto-create topics.
It then sends a message with a KafkaNull payload.
It then dynamically creates a new inbound adapter for a different topic si.new.topic and sends 10 messages there.
The message-driven adapter receives the messages and places them in a QueueChannel which the application reads and
writes to stdout:
Sending 10 messages...
Sending a null message...
GenericMessage [payload=foo0, headers={kafka_offset=657, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.topic}]
GenericMessage [payload=foo1, headers={kafka_offset=658, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.topic}]
GenericMessage [payload=foo2, headers={kafka_offset=659, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.topic}]
GenericMessage [payload=foo3, headers={kafka_offset=660, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.topic}]
GenericMessage [payload=foo4, headers={kafka_offset=661, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.topic}]
GenericMessage [payload=foo5, headers={kafka_offset=662, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.topic}]
GenericMessage [payload=foo6, headers={kafka_offset=663, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.topic}]
GenericMessage [payload=foo7, headers={kafka_offset=664, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.topic}]
GenericMessage [payload=foo8, headers={kafka_offset=665, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.topic}]
GenericMessage [payload=org.springframework.kafka.support.KafkaNull@4d9d1b69, headers={kafka_offset=667, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.topic}]
Adding an adapter for a second topic and sending 10 messages...
GenericMessage [payload=foo0, headers={kafka_offset=70, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.new.topic}]
GenericMessage [payload=foo1, headers={kafka_offset=71, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.new.topic}]
GenericMessage [payload=foo2, headers={kafka_offset=72, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.new.topic}]
GenericMessage [payload=foo3, headers={kafka_offset=73, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.new.topic}]
GenericMessage [payload=foo4, headers={kafka_offset=74, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.new.topic}]
GenericMessage [payload=foo5, headers={kafka_offset=75, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.new.topic}]
GenericMessage [payload=foo6, headers={kafka_offset=76, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.new.topic}]
GenericMessage [payload=foo7, headers={kafka_offset=77, kafka_receivedMessageKey=si.key, kafka_receivedPartitionId=0, kafka_receivedTopic=si.new.topic}]
Notice that the offset header increases on each run (the topic is not removed, to demonstrate that the offset is retained between executions).