reactor-kafka小试牛刀

序

本文主要展示一下如何使用reactor-kafka

maven

        
            io.projectreactor.kafka
            reactor-kafka
            1.0.1.RELEASE
        

准备

• 启动zookeeper

cd zookeeper-3.4.13
sh bin/zkServer.sh start
ZooKeeper JMX enabled by default
ZooKeeper remote JMX Port set to 8999
ZooKeeper remote JMX authenticate set to false
ZooKeeper remote JMX ssl set to false
ZooKeeper remote JMX log4j set to true
Using config: zookeeper-3.4.13/bin/../conf/zoo.cfg
-n Starting zookeeper ...
STARTED

• 启动kafka

cd kafka_2.11-1.1.1
sh bin/kafka-server-start.sh config/server.properties

• 创建topic

cd kafka_2.11-1.1.1
sh bin/kafka-topics.sh --create --topic demotopic --replication-factor 1 --partitions 3 --zookeeper localhost:2181
Created topic "demotopic".

实例

• producer

    @Test
    public void testProducer() throws InterruptedException {
        Map props = new HashMap<>();
        props.put(ProducerConfig.BOOTSTRAP_SERVERS_CONFIG, BOOTSTRAP_SERVERS);
        props.put(ProducerConfig.CLIENT_ID_CONFIG, "sample-producer");
        props.put(ProducerConfig.ACKS_CONFIG, "all");
        props.put(ProducerConfig.KEY_SERIALIZER_CLASS_CONFIG, IntegerSerializer.class);
        props.put(ProducerConfig.VALUE_SERIALIZER_CLASS_CONFIG, StringSerializer.class);
        SenderOptions senderOptions = SenderOptions.create(props);

        KafkaSender sender = KafkaSender.create(senderOptions);
        SimpleDateFormat dateFormat = new SimpleDateFormat("HH:mm:ss:SSS z dd MMM yyyy");

        CountDownLatch latch = new CountDownLatch(100);
        sender.send(Flux.range(1, 100)
                .map(i -> SenderRecord.create(new ProducerRecord<>(TOPIC, i, "Message_" + i), i)))
                .doOnError(e -> log.error("Send failed", e))
                .subscribe(r -> {
                    RecordMetadata metadata = r.recordMetadata();
                    System.out.printf("Message %d sent successfully, topic-partition=%s-%d offset=%d timestamp=%s\n",
                            r.correlationMetadata(),
                            metadata.topic(),
                            metadata.partition(),
                            metadata.offset(),
                            dateFormat.format(new Date(metadata.timestamp())));
                    latch.countDown();
                });

        latch.await(10, TimeUnit.SECONDS);
        sender.close();
    }

• consumer

    @Test
    public void testConsumer() throws InterruptedException {
        Map props = new HashMap<>();
        props.put(ConsumerConfig.BOOTSTRAP_SERVERS_CONFIG, BOOTSTRAP_SERVERS);
        props.put(ConsumerConfig.CLIENT_ID_CONFIG, "sample-consumer");
        props.put(ConsumerConfig.GROUP_ID_CONFIG, "sample-group");
        props.put(ConsumerConfig.KEY_DESERIALIZER_CLASS_CONFIG, IntegerDeserializer.class);
        props.put(ConsumerConfig.VALUE_DESERIALIZER_CLASS_CONFIG, StringDeserializer.class);
        props.put(ConsumerConfig.AUTO_OFFSET_RESET_CONFIG, "earliest");
        ReceiverOptions receiverOptions = ReceiverOptions.create(props);
        SimpleDateFormat dateFormat = new SimpleDateFormat("HH:mm:ss:SSS z dd MMM yyyy");

        CountDownLatch latch = new CountDownLatch(100);

        ReceiverOptions options = receiverOptions.subscription(Collections.singleton(TOPIC))
                .addAssignListener(partitions -> log.debug("onPartitionsAssigned {}", partitions))
                .addRevokeListener(partitions -> log.debug("onPartitionsRevoked {}", partitions));
        Flux> kafkaFlux = KafkaReceiver.create(options).receive();
        Disposable disposable = kafkaFlux.subscribe(record -> {
            ReceiverOffset offset = record.receiverOffset();
            System.out.printf("Received message: topic-partition=%s offset=%d timestamp=%s key=%d value=%s\n",
                    offset.topicPartition(),
                    offset.offset(),
                    dateFormat.format(new Date(record.timestamp())),
                    record.key(),
                    record.value());
            offset.acknowledge();
            latch.countDown();
        });

        latch.await(10, TimeUnit.SECONDS);
        disposable.dispose();
    }

小结

reactor-kafka对kafka的api进行封装，改造为reactive streams模式，这样用起来更为顺手，熟悉reactor的开发人员可以轻车熟路。

doc

• reactor-kafka-samples
• Reactor Kafka Reference Guide