21、Kafka 实战 - 消费者的配置细节

消费者的配置细节

配置

我们可以指定多个主题,分区,偏移量和消费者的并发数

@KafkaListener(groupId = "testGroup",  topicPartitions = {

 
        //@TopicPartition(topic = "test", partitions = { "0", "1" }),
        @TopicPartition(topic = "test", partitions = "0",
                partitionOffsets = @PartitionOffset(partition = "1", initialOffset = "100"))
}, concurrency = "3") //concurrency就是同组下的消费者个数,就是并发消费数,建议⼩于等于分区总数
public void listenGroup(ConsumerRecord < String, String > record,
                        Acknowledgment ack) {

 
    String value = record.value();
    System.out.println(value);
    System.out.println(record);
    //⼿动提交offset
    ack.acknowledge();
}

测试

我们启动项目后,消费者从test主题的1号分区的offset为100的地方开始消费