Spring For Apache Kafka 2.1.0 和 1.3.2 发布

Spring for Apache Kafka 2.1.0 已发布，同时发布的还有 1.3.2 和 2.0.2 维护版本，包含重要的 Bug 修复。

2.1.0 版本的主要将 kafka-clients 库升级到 1.0.0，以及一些改进：

• Sometimes, when a message can’t be processed, you may wish to stop the container so the condition can be corrected and the message re-delivered. The framework now provides the ContainerStoppingErrorHandler for record listeners and ContainerStoppingBatchErrorHandler for batch listeners.
• The KafkaAdmin now supports increasing partitions when a NewTopic bean is detected with a larger number of partitions than currently exist on the topic.
• StringJsonMessageConverter and JsonSerializer/JsonDeserializer now pass and consume type information in Headers. This allows multiple types to be easily sent/received on the same topic:

@SpringBootApplication
public class Kafka21Application {

    public static void main(String[] args) {
        SpringApplication.run(Kafka21Application.class, args)
            .close();
    }

    @Bean
    public ApplicationRunner runner(KafkaTemplate<Object, Object> template) {
        return args -> {
            template.send(MessageBuilder.withPayload(42)
                    .setHeader(KafkaHeaders.TOPIC, "blog")
                    .build());
            template.send(MessageBuilder.withPayload("43")
                    .setHeader(KafkaHeaders.TOPIC, "blog")
                    .build());
            Thread.sleep(5_000);
        };
    }

    @Bean
    public StringJsonMessageConverter converter() {
        return new StringJsonMessageConverter();
    }

    @Component
    @KafkaListener(id = "multi", topics = "blog")
    public static class Listener {

        @KafkaHandler
        public void intListener(Integer in) {
            System.out.println("Got an int: " + in);
        }

        @KafkaHandler
        public void stringListener(String in) {
            System.out.println("Got a string: " + in);
        }

    }

}

Got an int: 42
Got a string: 43

•  the JsonSerializer and JsonDeserializer can be configured using kafka properties for the producer/consumer.

更多详情请查阅发行主页。

转自 http://www.oschina.net/news/91110/spring-for-apache-kafka-2-1-0