You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user-zh@flink.apache.org by hahaha sc <sh...@gmail.com> on 2019/12/02 07:19:53 UTC

连续keyBy后sum

以下代码:

val env = StreamExecutionEnvironment.getExecutionEnvironment

val list = new ListBuffer[Tuple3[String,Int,Int]]

val random = new Random()

for(x<- 0 to 100){
  if(random.nextBoolean()){list.append(("INSERT",2,1))}else{
    list.append(("UPDATE",2,1))
  }
}


val data = env.fromElements(list).flatMap(_.toList)


val keyed=data.keyBy(0).sum(1)

keyed.print()


val reKeyed=keyed.keyBy(0).sum(2)
reKeyed.print()


按理说,reKeyed应该是在keyed的基础上,再进行相加,但是print出来的结果,reKeyed是在原始数据上进行的操作。

这是为什么呢?