Flink 异步IO

版权声明：本文为博主原创文章，遵循 CC 4.0 BY-SA 版权协议，转载请附上原文出处链接和本声明。

本文链接： https://blog.csdn.net/qq_22222499/article/details/92396131

Flink 异步IO

在flink我们经常需要与外部系统打交道,由于外部系统的问题，可能导致时间耗时比较长，为了不影响flink的处理性能，flink引入了异步IO来处理这个问题

实现

需要 extends RichAsyncFunction<IN, OUT>

分析

AsyncDataStream 有两个重要的方法

返回的结果可能是乱序的
public static <IN, OUT> SingleOutputStreamOperator<OUT> unorderedWait(
			DataStream<IN> in,
			AsyncFunction<IN, OUT> func,
			long timeout,
			TimeUnit timeUnit,
			int capacity //异常操作最大个数) {
		return addOperator(in, func, timeUnit.toMillis(timeout), capacity, OutputMode.UNORDERED);
	}
返回结果是有序的	
public static <IN, OUT> SingleOutputStreamOperator<OUT> orderedWait(
			DataStream<IN> in,
			AsyncFunction<IN, OUT> func,
			long timeout,
			TimeUnit timeUnit,
			int capacity) {
		return addOperator(in, func, timeUnit.toMillis(timeout), capacity, OutputMode.ORDERED);
	}	

demo

public class FlinkAsIoTest {

    public static void main(String[] args) {
        StreamExecutionEnvironment en = StreamExecutionEnvironment.getExecutionEnvironment();
        //自己实现的
        KafkaUtil util = new KafkaUtil();
        FlinkKafkaConsumer011 consumer1 = util.getConsumer("stream1", "test2");
        DataStream<String> source = en.addSource(consumer1);
        try {
            source.print();
            DataStream<String> result = AsyncDataStream.orderedWait(source, new AsIo(), 5000, TimeUnit.MILLISECONDS, 10);
            result.print();
            en.execute("test io");
        } catch (Exception e) {
            e.printStackTrace();
        }
    }
}

class AsIo extends RichAsyncFunction<String, String> {

    @Override
    public void asyncInvoke(String s, ResultFuture<String> resultFuture) throws Exception {

        CompletableFuture.supplyAsync(new Supplier<String>() {
            @Override
            public String get() {
                return queryResurlt(s);
            }
        }).thenAccept(new Consumer<String>() {
            @Override
            public void accept(String s) {
                resultFuture.complete(Collections.singleton(s));
            }
        });

    }


    /**
     * 假定这是一个异步操作
     *
     * @param params
     * @return
     */
    public String queryResurlt(String params) {
        try {
            //sleep 4s
            Thread.sleep(4000);
        } catch (InterruptedException e) {
            e.printStackTrace();
        }
        if (params.equals("beijing")) {
            return "this is beijing";
        } else if (params.equals("chendou")) {
            return "this is chendou";
        } else {
            return "this is other";
        }


    }

    @Override
    public void timeout(String input, ResultFuture<String> resultFuture) throws Exception {

    }
}