Spark 流（微批）处理（4）

Spark Streaming

def main(args: Array[String]): Unit = {
    // 1. 创建 Context
    val conf = new SparkConf()
      .setAppName("updateStateBykey")
      .setMaster("local[6]")
    val ssc = new StreamingContext(conf, Seconds(1))
    ssc.sparkContext.setLogLevel("ERROR")

    // 2. 读取tcp数据生成 DStream，
    val source = ssc.socketTextStream(
      hostname = "192.168.0.242",
      port = 9998,
      storageLevel = StorageLevel.MEMORY_AND_DISK_SER_2
    )

    // 3. 词频统计
    val wordsTuple = source.flatMap(_.split(" "))
      .map((_, 1))

    // 4. 全局聚合,累计每个批次得结果
    ssc.checkpoint("checkpoint")
    def updateFunc(newValue: Seq[Int], runningValue: Option[Int]): Option[Int] = {
      // newValue : 对应当前批次中 Key 对应的所有 Value
      // runningValue : 当前的中间结果
      val currBatchValue = newValue.sum
      val state = runningValue.getOrElse(0) + currBatchValue
      Some(state)
    }
    val result = wordsTuple.updateStateByKey[Int](updateFunc _)
    
    // 5. 输出
    result.print()
    ssc.start()
    ssc.awaitTermination()
  }

/**
     * -------------------------------------------
     * Time: 1603871685000 ms
     * -------------------------------------------
     * (hadoop,1)
     * (hello,1)
     *
     * -------------------------------------------
     * Time: 1603871686000 ms
     * -------------------------------------------
     * (hadoop,1)
     * (hello,2)
     * (spark,1)
     * 
     * -------------------------------------------
     * Time: 1603871692000 ms
     * -------------------------------------------
     * (hadoop,1)
     * (ok,1)
     * (hello,2)
     * (spark,2)
     */

Structured Streaming

最后

以上就是幸福路灯最近收集整理的关于Spark 流（微批）处理（4）的全部内容，更多相关Spark内容请搜索靠谱客的其他文章。