6. kafka拦截器

• 拦截器定义
  拦截器参数命名为：interceptor.classes。官方文档解析如下：

A list of classes to use as interceptors. Implementing the org.apache.kafka.clients.producer.ProducerInterceptor interface allows you to intercept (and possibly mutate) the records received by the producer before they are published to the Kafka cluster. By default, there are no interceptors.

即拦截发送到kafka服务器之前的消息，且在序列化&反序列化之前调用，序列化&反序列化又在分区策略之前调用，这个调用顺序在[kafka生产者&消费者]已经分析过了。拦截器允许修改key和value，同时拦截器可以指定多个，多个拦截器形成拦截器链，且有先后顺序，假定按照如下方式配置2个拦截器，那么会先调用AProducerInterceptor，再调用BProducerInterceptor，且调用BProducerInterceptor时的ProducerRecord是经过AProducerInterceptor修改过的ProducerRecord（如果在AProducerInterceptor中修改过ProducerRecord的话）：

props.put("interceptor.classes", "com.afei.kafka.interceptor.AProducerInterceptor,com.afei.kafka.interceptor.BProducerInterceptor");

• ProducerInterceptor接口定义源码解读

public interface ProducerInterceptor extends Configurable {
    /**
     * This is called from {@link org.apache.kafka.clients.producer.KafkaProducer#send(ProducerRecord)} and
     * {@link org.apache.kafka.clients.producer.KafkaProducer#send(ProducerRecord, Callback)} methods, before key and value
     * get serialized and partition is assigned (if partition is not specified in ProducerRecord).
     * 即这个方法调用在KafkaProducer.send()之后，但是在key&value序列化之前，以及分配分区之前。
     * 

     * This method is allowed to modify the record, in which case, the new record will be returned. The implication of modifying
     * key/value is that partition assignment (if not specified in ProducerRecord) will be done based on modified key/value,
     * 这个方法中允许修改ProducerRecord即发送的消息，接下来的分区分配，根据修改后的key（如果会在onSend()中修改key的话）进行计算.
     */
    public ProducerRecord onSend(ProducerRecord record);

    /**
     * 不管消息发送成功还是发送过程抛出异常，这个方法都会执行
     */
    public void onAcknowledgement(RecordMetadata metadata, Exception exception);

    /**
     * This is called when interceptor is closed
     */
    public void close();
}

• 自定义拦截器实现

/**
 * @author afei
 * @version 1.0.0
 * @since 2018年06月27日
 */
public class MyProducerInterceptor implements ProducerInterceptor {

    /**
     * 统计发送成功数
     */
    private static AtomicLong sendSuccess = new AtomicLong(0);
    /**
     * 统计发送失败数
     */
    private static AtomicLong sendFailure = new AtomicLong(0);

    @Override
    public ProducerRecord onSend(ProducerRecord record) {

        this.outputSendStat();

        // 改写ProducerRecord, 将key置为null, 分区全部交给kafka去决定
        return new ProducerRecord<>(record.topic(),
                record.partition(), record.timestamp(), null, record.value(),
                record.headers());
    }

    @Override
    public void onAcknowledgement(RecordMetadata metadata, Exception exception) {
        // 如果没有异常表示发送成功, 那么发送成功数+1, 否则发送失败数+1
        if (exception!=null){
            sendFailure.getAndIncrement();
        }else{
            sendSuccess.getAndIncrement();
        }
    }

    /**
     * 打印出发送的成功&失败次数的统计信息
     */
    private void outputSendStat(){
        long successCount = sendSuccess.get();
        long  failedCount = sendFailure.get();
        System.out.println("success count: "+successCount+", failed count:"+failedCount);
    }

    @Override
    public void close() {
        this.outputSendStat();
    }

    @Override
    public void configure(Map configs) {

    }
}