Linux Kafka 3.5 KRaft模式集群部署

这里是weihubeats,觉得文章不错可以关注公众号小奏技术，文章首发。拒绝营销号，拒绝标题党

背景

kafka在KIP-500引入了KRaft替代Zookeeper来实现自我管理元数据

详细信息可以看原文链接

• KIP-500

KRaft简介

KRaft是kafka用来取代zookeeper的分布式协调管理组件。

架构改变

原先依赖于Zookeeper选举出一个controller
现在由KRaft集群中自己选举，产生一个controller

优点

• Kafka不用再依赖外部框架，能够做到独立运行
• Kafka集群扩展时不用再受到Zookeeper读写能力的限制

更多优点和缺点这里暂时不太多讨论主要以部署为主

部署3节点kafaka集群

KRaft部署方式支持controller和broker在同一进程。也支持分开部署
线上推荐分开部署。这里由于是测试集群，打算controller和broker在同一进程部署

记得所有机器9092、9093端口打开

下载Kafka

wget https://dlcdn.apache.org/kafka/3.5.0/kafka_2.13-3.5.0.tgz

这里第一次下载报错说证书已过期,添加证书忽略下载

wget --no-check-certificate https://dlcdn.apache.org/kafka/3.5.0/kafka_2.13-3.5.0.tgz

发现国内服务器下载国外软件还是非常慢。最终决定找国内镜像。

• 阿里云Kafka镜像:http://mirrors.aliyun.com/apache/kafka/3.5.0/?spm=a2c6h.25603864.0.0.3c7d126emg02YS

使用国内镜像下载

wget http://mirrors.aliyun.com/apache/kafka/3.5.0/kafka_2.13-3.5.0.tgz

三台机器都执行

解压

tar -xzf kafka_2.13-3.5.0.tgz

三台机器都执行

给集群生成一个UUID

我们进入到解压的bin目录，我这里是/data/kafka_2.13-3.5.0/bin
然后执行如下命令

kafka_2.13-3.0.0/bin/kafka-storage.sh random-uuid

单台机器生成即可

执行完会生产一个字符串，类似这样xgK3spReSO7ijVK4rEbbbQ

格式化存储路径

sh kafka-storage.sh format -t xgK3spReSO7ijVK4rEbbbQ  -c ../config/kraft/server.properties

三台机器都执行

修改配置

我们这里要修改三台机器的server.properties配置
我这里的路径是在/data/kafka_2.13-3.5.0/config/kraft/server.properties

• node1

node.id = 1
controller.quorum.voters = 1@192.168.1.1:9093,2@92.168.1.2:9093,3@92.168.1.3:9093
process.roles = broker,controller
listeners=PLAINTEXT://192.168.1.1:9092,CONTROLLER://92.168.1.1:9093
log.dirs=/data/kakfa01/logs

• node2

node.id = 2
controller.quorum.voters = 1@192.168.1.1:9093,2@92.168.1.2:9093,3@92.168.1.3:9093
process.roles = broker,controller
listeners=PLAINTEXT://192.168.1.2:9092,CONTROLLER://92.168.1.2:9093
log.dirs=/data/kakfa02/logs

• node3

node.id = 3
controller.quorum.voters = 1@192.168.1.1:9093,2@92.168.1.2:9093,3@92.168.1.3:9093
process.roles = broker,controller
listeners=PLAINTEXT://192.168.1.3:9092,CONTROLLER://92.168.1.3:9093
log.dirs=/data/kakfa03/logs

启动集群

export KAFKA_HEAP_OPTS="-Xmx4G -Xms4G"&&nohup sh /data/kafka_2.13-3.5.0/bin/kafka-server-start.sh /data/kafka_2.13-3.5.0/config/kraft/server.properties &

三台机器都执行

启动完我们就有了一个三节点的kafka集群

测试

创建topic

sh kafka-topics.sh --create --topic xiaozou --partitions 1 --replication-factor 1 --bootstrap-server 192.168.1.1:9092,192.168.1.2:9092,192.168.1.3:9092

查看topic

sh kafka-topics.sh --list --bootstrap-server 192.168.1.1:9092,192.168.1.2:9092,192.168.1.3:9092

代码测试

• 生产消息

public class KafkaProducer {

    private static final String TOPIC = "xiaozou";
    private static final String BOOTSTRAP_SERVERS = "192.168.1.1:9092,192.168.1.2:9092,192.168.1.3:9092";

    public static void main(String[] args) {
        // 生产消息
        produceMessage();
    }

    private static void produceMessage() {
        Properties props = new Properties();
        props.put(ProducerConfig.BOOTSTRAP_SERVERS_CONFIG, BOOTSTRAP_SERVERS);
        props.put(ProducerConfig.KEY_SERIALIZER_CLASS_CONFIG, IntegerSerializer.class);
        props.put(ProducerConfig.VALUE_SERIALIZER_CLASS_CONFIG, StringSerializer.class);

        Producer<String, String> producer = new org.apache.kafka.clients.producer.KafkaProducer<>(props);

        try {
            for (int i = 0; i < 10; i++) {
                String message = "小奏message " + i;
                System.out.println("开始发送消息");
                Future<RecordMetadata> send = producer.send(new ProducerRecord<>(TOPIC, message));
                RecordMetadata recordMetadata = send.get();
                System.out.println("Produced message: " + message);
            }
        } catch (Exception e) {
            e.printStackTrace();
        } finally {
            producer.close();
        }
    }

}

• 消费消息

public class KafkaConsumerExample {

    private static final String TOPIC_NAME = "xiaozou";

    private static final String GROUP_ID = "xiaozou_gid";

    private static final String BOOTSTRAP_SERVERS = "192.168.1.1:9092,192.168.1.2:9092,192.168.1.3:9092";


    public static void main(String[] args) {
        Properties props = new Properties();
        props.put(ConsumerConfig.BOOTSTRAP_SERVERS_CONFIG, BOOTSTRAP_SERVERS);
        props.put(ConsumerConfig.GROUP_ID_CONFIG, GROUP_ID);
        props.put(ConsumerConfig.KEY_DESERIALIZER_CLASS_CONFIG, IntegerDeserializer.class);
        props.put(ConsumerConfig.VALUE_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class);



        // 创建消费者实例
        KafkaConsumer<String, String> consumer = new KafkaConsumer<>(props);
        // 订阅主题
        consumer.subscribe(Collections.singletonList(TOPIC_NAME));

        // 消费消息
        try {
            while (true) {
                ConsumerRecords<String, String> records = consumer.poll(100);
                for (ConsumerRecord<String, String> record : records) {
                    System.out.println("接收到消息：key = " + record.key() + ", value = " + record.value() +
                        ", partition = " + record.partition() + ", offset = " + record.offset());
                }
                consumer.commitSync(); // 手动提交偏移量
            }
        } finally {
            consumer.close();
        }
    }
}