Flink AggregatingState 实例

AggregatingState介绍

• AggregatingState需要和AggregateFunction配合使用
• add()方法添加一个元素，触发AggregateFunction计算
• get()获取State的值

需求：计算每个设备10秒内的平均温度

import org.apache.flink.api.common.eventtime.SerializableTimestampAssigner;
import org.apache.flink.api.common.eventtime.WatermarkStrategy;
import org.apache.flink.api.common.functions.AggregateFunction;
import org.apache.flink.api.common.state.AggregatingState;
import org.apache.flink.api.common.state.AggregatingStateDescriptor;
import org.apache.flink.api.common.typeinfo.TypeHint;
import org.apache.flink.api.common.typeinfo.TypeInformation;
import org.apache.flink.api.java.functions.KeySelector;
import org.apache.flink.api.java.tuple.Tuple2;
import org.apache.flink.api.java.tuple.Tuple3;
import org.apache.flink.configuration.Configuration;
import org.apache.flink.streaming.api.datastream.DataStreamSource;
import org.apache.flink.streaming.api.environment.StreamExecutionEnvironment;
import org.apache.flink.streaming.api.functions.KeyedProcessFunction;
import org.apache.flink.streaming.api.functions.source.SourceFunction;
import org.apache.flink.util.Collector;

import java.time.Duration;
import java.util.Random;

public class AggregatingStateTest {
    public static void main(String[] args) throws Exception {
        // 计算每个设备10s内温度的平均值
        StreamExecutionEnvironment env = StreamExecutionEnvironment.getExecutionEnvironment();
        env.getConfig().setAutoWatermarkInterval(100l);

        DataStreamSource> tuple3DataStreamSource = env.addSource(new SourceFunction>() {
            boolean flag = true;

            @Override
            public void run(SourceContext> ctx) throws Exception {
                String[] str = {"水阀1", "水阀2", "水阀3"};
                while (flag) {
                    int i = new Random().nextInt(3);
                    // 温度
                    int temperature = new Random().nextInt(100);
                    Thread.sleep(1000l);
                    // 设备号、温度、事件时间
                    ctx.collect(new Tuple3(str[i], temperature, System.currentTimeMillis()));
                }
            }

            @Override
            public void cancel() {
                flag = false;
            }
        });

        tuple3DataStreamSource.assignTimestampsAndWatermarks(WatermarkStrategy.>forBoundedOutOfOrderness(Duration.ofSeconds(2))
                .withTimestampAssigner(new SerializableTimestampAssigner>() {
                    @Override
                    public long extractTimestamp(Tuple3 stringIntegerLongTuple3, long l) {
                        return stringIntegerLongTuple3.f2;
                    }
                })).keyBy(new KeySelector, String>() {
            @Override
            public String getKey(Tuple3 stringIntegerLongTuple3) throws Exception {
                return stringIntegerLongTuple3.f0;
            }
        }).process(new KeyedProcessFunction, String>() {
            Long interval = 10 * 1000l;
            // 这个类型是aggregatingState中的输入和输出类型
            AggregatingState aggregatingState = null;
            @Override
            public void open(Configuration parameters) throws Exception {
                super.open(parameters);
                // , Double>这是输入，中间状态，输出类型。TypeInformation.of(new TypeHint>(){})这个是aggregatingState存储的数据的类型
                AggregatingStateDescriptor, Double> aggregatingStateDescriptor =
                        new AggregatingStateDescriptor, Double>("aggregatingState", new MyAggregate(), TypeInformation.of(new TypeHint>(){}));
                aggregatingState = getRuntimeContext().getAggregatingState(aggregatingStateDescriptor);
            }

            @Override
            public void processElement(Tuple3 value, Context ctx, Collector out) throws Exception {
                // 10s的起始的时间
                Long start = ctx.timestamp() - (ctx.timestamp() % interval);
                Long timerTimestamp = start + interval;
                ctx.timerService().registerEventTimeTimer(timerTimestamp);
                aggregatingState.add(value.f1);
            }

            @Override
            public void onTimer(long timestamp, OnTimerContext ctx, Collector out) throws Exception {
                super.onTimer(timestamp, ctx, out);
                Double aDouble = aggregatingState.get();
                String str = "[" + ctx.getCurrentKey() + "] " + "十秒内的平均温度为：" + aDouble;
                out.collect(str);
            }
        }).print();

        env.execute("aggregatingState");
    }

    private static class MyAggregate implements AggregateFunction, Double> {

        @Override
        public Tuple2 createAccumulator() {
            // 初始化温度和次数
            return new Tuple2(0,0);
        }

        @Override
        public Tuple2 add(Integer integer, Tuple2 integerIntegerTuple2) {
            // 历史温度加上本次温度，次数加1
            return new Tuple2(integerIntegerTuple2.f0 + integer, integerIntegerTuple2.f1 +1);
        }

        @Override
        public Double getResult(Tuple2 integerIntegerTuple2) {
            return Double.valueOf(integerIntegerTuple2.f0 / integerIntegerTuple2.f1);
        }

        @Override
        public Tuple2 merge(Tuple2 integerIntegerTuple2, Tuple2 acc1) {
            return new Tuple2(integerIntegerTuple2.f0 + acc1.f0, integerIntegerTuple2.f1 + acc1.f1);
        }
    }
}