Kafka生产者之自定义分区器

1. 没有设置key

我们的消息会被轮训的发送到不同的分区

2. 设置了key

kafka自带的分区器，会根据key计算出来一个hash值，这个hash值会对应某一个分区

如果key相同的，那么hash值必然相同，key相同的值，必然是会被发送到同一个分区

3. 自定义分区器

package com.bigdata.test;

import org.apache.kafka.clients.producer.Partitioner;
import org.apache.kafka.common.Cluster;
import org.apache.kafka.common.PartitionInfo;

import java.util.List;
import java.util.Map;
import java.util.Random;

/**
 * Copyright (c) 2019 bigdata ALL Rights Reserved
 * Project: learning
 * Package: com.bigdata.test
 * Version: 1.0
 *
 * @author qingzhi.wu 2020/8/1 17:01
 */
public class MyPartitionor implements Partitioner {
    private Random random;
    public int partition(String topic, Object keyObj, byte[] keyBytes, Object value, byte[] valueBytes, Cluster cluster) {
        String key = (String) keyObj;
        //获取这个分区的所有信息
        List<PartitionInfo> partitionInfos = cluster.availablePartitionsForTopic(topic);
        //获取到分区个数
        int count = partitionInfos.size();
        //热点分区
        int hotDataPartition = count - 1;

        return !key.contains("hot_data") ? random.nextInt(count - 1):hotDataPartition;
    }

    public void close() {

    }

    public void configure(Map<String, ?> configs) {
        random = new Random();
    }
}