使用Flink DataStream计算窗口持续时间的平均值

Question

我正在使用Flink DataStream API,其中有机架可用,我想按机架ID计算温度组的"平均值".我的窗口持续时间为40秒,我的窗口每10秒钟滑动一次......以下是我的代码,我计算每个rackID每10秒的温度总和,但现在我想计算平均温度::

static Properties properties=new Properties();
    public static Properties getProperties()
    {
        properties.setProperty("bootstrap.servers", "54.164.200.104:9092");
        properties.setProperty("zookeeper.connect", "54.164.200.104:2181");
        //properties.setProperty("deserializer.class", "kafka.serializer.StringEncoder");
        //properties.setProperty("group.id", "akshay");
        properties.setProperty("auto.offset.reset", "earliest");
        return properties;
    }

 @SuppressWarnings("rawtypes")
public static void main(String[] args) throws Exception 
{
    StreamExecutionEnvironment env=StreamExecutionEnvironment.getExecutionEnvironment();
    env.setStreamTimeCharacteristic(TimeCharacteristic.EventTime);
    Properties props=Program.getProperties();
    DataStream<TemperatureEvent> dstream=env.addSource(new FlinkKafkaConsumer09<TemperatureEvent>("TemperatureEvent", new TemperatureEventSchema(), props)).assignTimestampsAndWatermarks(new IngestionTimeExtractor<>());
    DataStream<TemperatureEvent> ds1=dstream.keyBy("rackId").timeWindow(Time.seconds(40), Time.seconds(10)).sum("temperature");
    env.execute("Temperature Consumer");
}

如何计算上述例子的平均温度？

Answer 1

据我所知，您需要自己编写平均值函数。您可以在此处找到示例：

https://github.com/apache/flink/blob/master/flink-examples/flink-examples-streaming/src/main/java/org/apache/flink/streaming/examples/windowing/GroupedProcessingTimeWindowExample.java

就您而言，您可能会替换 .sum("temperature");

使用类似的东西 .apply(new Avg()); 并实现Avg类：

public class Avg implements WindowFunction<TemperatureEvent,  TemperatureEvent, Long, org.apache.flink.streaming.api.windowing.windows.Window> {

  @Override
  public void apply(Long key, Window window, Iterable<TemperatureEvent> values, Collector<TemperatureEvent> out) {
    long sum = 0L;
    int count = 0;
    for (TemperatureEvent value : values) {
        sum += value.getTemperature();
        count ++;
    }

    TemperatureEvent result = values.iterator().next();
    result.setTemperature(sum / count);
    out.collect(result);
  }
}

注意：如果有可能在空窗口上调用您的函数（例如通过使用自定义触发器），则在访问elements.head之前需要检查