1、启动kafka和flink以及hadoop

为了方便操作，mac可以下载Tabby软件

1.1 启动kafka：进入kafka目录
	bin/kafka-server-start.sh config/server.properties
1.2 创建kafka
	进入bin目录   
	./kafka-topics.sh —create —bootstrap-server qf01:9092 —replication-factor 1 —partition 1 —topic test
1.3 启动生产者和消费者
    生产者：./kafka-console-producer.sh --broker-list qf01:9092 --topic test
    消费者：./kafka-console-consumer.sh --bootstrap-server qf01:9092 --topic test --from-beginning
1.4 启动hadoop集群
	start-all.sh
1.5 启动zk集群
	zkServer.sh start
1.6 启动flink
	start-cluster.sh

2、代码实现

2.1 创建生产者的代码

JProducer.class

@Configuration
@Slf4j
public class JProducer extends Thread {
    public static final String broker_list = "ip(换成自己的):9092";
    public static final String topic = "test";

    public static void main(String[] args) {
        JProducer jproducer = new JProducer();
        jproducer.start();
    }

    @Override
    public void run() {
        try {
            producer();
        } catch (InterruptedException e) {
            throw new RuntimeException(e);
        }
    }


    /**
     * 向kafka批量生成记录
     */
    @Scheduled(initialDelayString="${kf.flink.init}",fixedDelayString = "${kf.flink.fixRate}")
    private void producer() throws InterruptedException {
        log.info("启动定时任务");
        Properties props = config();//kafka连接
        Producer<String, String> producer = new KafkaProducer<>(props);
        Date date = new Date();
        SimpleDateFormat simpleDateFormat = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss");
        String dateString = simpleDateFormat.format(date);
        while (true) {
            for (int i = 1; i <= Integer.MAX_VALUE; i++) {
                String json = "{\"id\":" + i + ",\"ip\":\"192.168.0." + i + "\",\"date\":" + dateString + "}";
                String k = "第" + i + "条数据=" + json;
                sleep(300);
                if (i % 10 == 0) {
                    sleep(1000);
                }
                producer.send(new ProducerRecord<String, String>(topic, k, json));
            }
           producer.close();
        }
    }


    /**
     * kafka连接
     * @return
     */
    private Properties config() {
        Properties props = new Properties();
        props.put("bootstrap.servers",broker_list);
        props.put("acks", "1");
        props.put("retries", 0);
        props.put("batch.size", 16384);
        props.put("linger.ms", 1);
        props.put("buffer.memory", 33554432);
        props.put("key.serializer", "org.apache.kafka.common.serialization.StringSerializer");
        props.put("value.serializer", "org.apache.kafka.common.serialization.StringSerializer");
        return props;
    }


}

2.2 创建flink程序

Flink.class

public class Flink {
    private static final String topic = "test";
    public static final String broker_list = "ip(换成自己的):9092";

    public static void main(String[] args) {
        final StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        env.enableCheckpointing(1000);
        env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime);
        //读取Kafka数据，主题topic:
        DataStream<String> transction = env.addSource(new FlinkKafkaConsumer<String>(topic, new SimpleStringSchema(), props()).setStartFromEarliest());
        transction.rebalance().map(new MapFunction<String, Object>() {
            private static final long serialVersionUID = 1L;
            @Override
            public String map(String value) {
                System.out.println("ok了");
                return value;
            }
        }).print();

        try {
            env.execute();
        } catch (Exception ex) {
            ex.printStackTrace();
        }


    }

    public static Properties props() {
        Properties props = new Properties();
        props.put("bootstrap.servers", broker_list);
        props.put("zookeeper.connect", "192.168.47.130:2182");
        props.put("group.id", "kv_flink");
        props.put("enable.auto.commit", "true");
        props.put("auto.commit.interval.ms", "1000");
        props.put("key.deserializer", "org.apache.kafka.common.serialization.StringDeserializer");
        props.put("value.deserializer", "org.apache.kafka.common.serialization.StringDeserializer");
        return props;
    }

}

3、把项目打包放入hdfs中

hdfs dfs -put xxx.jar /root/xxx.jar

4、执行flink程序

	进入flink的bin目录
	./flink run --class com.xxx.xxx.MyWordCount1  /root/jar

5、打开web页面

6、额外扩展点

主要是配置一些属性字段，可以参考flink的的配置文件
到flink/conf/flink-conf.yaml中

      //一旦开启是耗性能的 这是创建一次 300ms创建一次
        env.enableCheckpointing(300);
        env.getCheckpointConfig().setCheckpointingMode(CheckpointingMode.EXACTLY_ONCE);
        env.getCheckpointConfig().setMaxConcurrentCheckpoints(2);
        //2、设置重启策略,重启三次，没10秒执行一次   failureRateRestart 失败率重启
        env.setRestartStrategy(RestartStrategies.fixedDelayRestart(3,10000L));
        //3、并行度setParallelism()
        env.setParallelism(1);
        //4、时间特性  参数 TimeCharacteristic 有三种类型： ProcessingTime, IngestionTime, EventTime;
        env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime);
        //5、并行任务调度策略
        //1、EagerSchedulingStrategy：适用于流计算，同时调度所有的task
        //2、LazyFromSourcesSchedulingStrategy：适用于批处理，当输入数据准备好时（上游处理完）进行vertices调度。
        //3、PipelinedRegionSchedulingStrategy：以流水线的局部为粒度进行调度。