kafka消息分发策略分析

当我们使用kafka向指定Topic发送消息时，如果该Topic具有多个partition，无论消费者有多少，最终都会保证一个partition内的消息只会被一个Consumer group中的一个Consumer消费，也就是说同一Consumer group中的多个Consumer自动会起到负载均衡的效果。

1、消息构造

下面我们就针对调用kafka API发送消息到Topic时partition的分配策略，分析下其内部具体的源码码实现。

首先看下kafka API中消息体ProducerRecord类的构造函数，可以看到构造消息时可指定该消息要发送的Topic、partition、key、value等关键信息。

    /**
     * Creates a record to be sent to a specified topic and partition
     *
     * @param topic The topic the record will be appended to
     * @param partition The partition to which the record should be sent
     * @param key The key that will be included in the record
     * @param value The record contents
     * @param headers The headers that will be included in the record
     */
    public ProducerRecord(String topic, Integer partition, K key, V value, Iterable<Header> headers) {
        this(topic, partition, null, key, value, headers);
    }

    /**
     * Creates a record to be sent to a specified topic and partition
     *
     * @param topic The topic the record will be appended to
     * @param partition The partition to which the record should be sent
     * @param key The key that will be included in the record
     * @param value The record contents
     */
    public ProducerRecord(String topic, Integer partition, K key, V value) {
        this(topic, partition, null, key, value, null);
    }

    /**
     * Create a record to be sent to Kafka
     *
     * @param topic The topic the record will be appended to
     * @param key The key that will be included in the record
     * @param value The record contents
     */
    public ProducerRecord(String topic, K key, V value) {
        this(topic, null, null, key, value, null);
    }

2、分发策略 

在实际使用中，我们一般不会指定消息发送的具体partition，最多只会传入key值，类似下面这种方式：

producer.send(new ProducerRecord<Object, Object>(topic, key, data));

而kafka也会根据你传入key的hash值，通过取余的方法，尽可能保证消息能够相对均匀的分摊到每个可用的partition上；

下面是kafka内部默认的分发策略：

public class DefaultPartitioner implements Partitioner {

    private final ConcurrentMap<String, AtomicInteger> topicCounterMap = new ConcurrentHashMap<>();

    public void configure(Map<String, ?> configs) {}

    /**
     * Compute the partition for the given record.
     *
     * @param topic The topic name
     * @param key The key to partition on (or null if no key)
     * @param keyBytes serialized key to partition on (or null if no key)
     * @param value The value to partition on or null
     * @param valueBytes serialized value to partition on or null
     * @param cluster The current cluster metadata
     */
    public int partition(String topic, Object key, byte[] keyBytes, Object value, byte[] valueBytes, Cluster cluster) {
        //获取该topic的分区列表
        List<PartitionInfo> partitions = cluster.partitionsForTopic(topic);
        int numPartitions = partitions.size();
        //如果key值为null
        if (keyBytes == null) {
            //维护一个key为topic的ConcurrentHashMap，并通过CAS操作的方式对value值执行递增+1操作
            int nextValue = nextValue(topic);
            //获取该topic的可用分区列表
            List<PartitionInfo> availablePartitions = cluster.availablePartitionsForTopic(topic);
            if (availablePartitions.size() > 0) {//如果可用分区大于0
                //执行求余操作，保证消息落在可用分区上
                int part = Utils.toPositive(nextValue) % availablePartitions.size();
                return availablePartitions.get(part).partition();
            } else {
                // 没有可用分区的话，就给出一个不可用分区
                return Utils.toPositive(nextValue) % numPartitions;
            }
        } else {
            // 通过计算key的hash，确定消息分区
            return Utils.toPositive(Utils.murmur2(keyBytes)) % numPartitions;
        }
    }

    private int nextValue(String topic) {
        //获取一个AtomicInteger对象
        AtomicInteger counter = topicCounterMap.get(topic);
        if (null == counter) {//如果为空
            //生成一个随机数
            counter = new AtomicInteger(ThreadLocalRandom.current().nextInt());
            //维护到topicCounterMap中
            AtomicInteger currentCounter = topicCounterMap.putIfAbsent(topic, counter);
            if (currentCounter != null) {
                counter = currentCounter;
            }
        }
        //返回值并执行递增
        return counter.getAndIncrement();
    }

    public void close() {}

}

3、自定义负载策略

我们也可以通过实现Partitioner接口，自定义分发策略，看下具体实现

自定义实现Partitioner接口

/**
 * 自定义实现Partitioner接口
 *
 */
public class KeyPartitioner implements Partitioner {

    /**
     * 实现具体分发策略
     */
    @Override
    public int partition(String topic, Object key, byte[] bytes, Object o1, byte[] bytes1, Cluster cluster) {
        List<PartitionInfo> availablePartitions = cluster.availablePartitionsForTopic(topic);//拉取可用的partition
        if (key == null||key.equals("")) {
            int random =  (int) (Math.random() * 10);
            int part = random % availablePartitions.size();
            return availablePartitions.get(part).partition();
        }
        return  Math.abs(key.toString().hashCode() % 6);
    }

    @Override
    public void configure(Map<String, ?> configs) {
        // TODO Auto-generated method stub

    }

    @Override
    public void close() {
        // TODO Auto-generated method stub

    }

}

同时在初始化kafka生产者时，增加自定义配置

Properties properties = new Properties();
properties.put(ProducerConfig.PARTITIONER_CLASS_CONFIG,KeyPartitioner.class); //加入自定义的配置
producer = new KafkaProducer<Object, Object>(properties);

 4、总结

以上是对kafka消息分发的策略进行一定的分析与自定义扩展，希望对大家在使用kafka时有所帮助，其中如有不足与不正确的地方还望指出与海涵。

关注微信公众号，查看更多技术文章。