基于Spark-Streaming滑动窗口实现——实时排名与统计

基于Spark-Streaming滑动窗口实现——实时排名与统计

1、主流程

 def main(args: Array[String]) {
    val conf = new SparkConf().setAppName("StockerRealRank"); //.setMaster("local[5]");
    val sc = new SparkContext(conf);
    val ssc = new StreamingContext(sc, Seconds(5));
    //缓存2天的数据
    ssc.remember(Minutes(60 * 48));
    val sqlContext = new HiveContext(sc);

    Logger.getLogger("org.apache.spark").setLevel(Level.WARN);
    Logger.getLogger("org.eclipse.jetty.server").setLevel(Level.ERROR);

    //1.注册UDF
    val udf = UDFUtils();
    udf.registerUdf(sqlContext);

    //2.kafka数据处理
    val kafkaService = KakfaService();
    val urlClickLogPairsDStream = kafkaService.kafkaDStreamForStocker(ssc);

    //3.缓存hive中的数据
    val cacheUtils = CacheUtils();
    cacheUtils.cacheStockInfoData(sqlContext);

    //4.缓存窗口函数数据处理
    val urlClickCountsDStream = urlClickLogPairsDStream.reduceByKeyAndWindow(
      (v1: Int, v2: Int) => {
        v1 + v2
      },
      Minutes(60 * 2),
      Seconds(25));

    //第二次消费kafka数据
    val urlClickCountsDStreamByDay = urlClickLogPairsDStream.reduceByKeyAndWindow(
      (v1: Int, v2: Int) => {
        v1 + v2
      },
      Minutes(60 * 48),
      Seconds(35));

    //5.处理业务逻辑
    urlClickCnt(urlClickCountsDStream, sqlContext);
    urlClickCntByDay(urlClickCountsDStreamByDay, sqlContext);

    //6.启动streaming任务
    ssc.start();
    ssc.awaitTermination();

  }

2、注册UDF，采用scala的伴生对象实现

import org.apache.spark.sql.hive.HiveContext
import java.util.regex.Pattern

/**
 * @author Administrator
 */
class UDFUtils {
  def registerUdf(sqlContext: HiveContext) {
    sqlContext.udf.register("strLen", (str: String) => str.length())

    sqlContext.udf.register("concat", (str1: String, str2: String, str3: String) => str1 + str2 + str3)

    sqlContext.udf.register("concat4", (str1: String, str2: String, str3: String, str4: String) => str1 + str2 + str3 + str4)

    sqlContext.udf.register("regexp_extract", (str: String, pattern: String) => {
      val matcher = Pattern.compile(pattern, 1).matcher(str)
      var res = ""
      while (matcher.find()) {
        res = matcher.group()
      }
      res
    })

    sqlContext.udf.register("getHost", (url: String) => {
      var strURL = "";
      try {
        strURL = url.toString();
        if (strURL.contains("://") && (strURL.indexOf("://") < 6) && strURL.length() > (strURL.indexOf("://") + 4)) {
          strURL = strURL.substring(strURL.indexOf("://") + 3);
        }

        if (strURL.contains("/")) {
          strURL = strURL.substring(0, strURL.indexOf("/"));
        }

        if (strURL.contains(":")) {
          strURL = strURL.substring(0, strURL.indexOf(":"));
        }

      } catch {
        case e: Exception => println("registerUdf Exception")
      }
      strURL;
    })

  }
}

object UDFUtils {
  def apply() = new UDFUtils();
}

3、kafka数据处理

import org.apache.spark.streaming.StreamingContext
import org.apache.spark.streaming.dstream.DStream
import org.apache.spark.streaming.kafka.KafkaUtils
import kafka.serializer.StringDecoder
import java.util.regex.Pattern

/**
 * @author Administrator
 */
class KakfaService {
  def kafkaDStream(ssc: StreamingContext): DStream[(String, Int)] = {
    val topics = Set("teststreaming");

    val brokers = "bdc46.hexun.com:9092,bdc53.hexun.com:9092,bdc54.hexun.com:9092";

    val kafkaParams = Map[String, String]("metadata.broker.list" -> brokers, "serializer.class" -> "kafka.serializer.StringEncoder");

    // Create a direct stream
    val kafkaStream = KafkaUtils.createDirectStrea