Spark Streaming（五）：与Spark SQL整合

Spark Streaming最强大的地方在于，可以与Spark Core、Spark SQL整合使用，之前已经通过transform、foreachRDD等算子看到，如何将DStream中的RDD使用Spark Core执行批处理操作。现在就来看看，如何将DStream中的RDD与Spark SQL结合起来使用。

Demo：每隔10秒，统计最近60秒的，每个种类的每个商品的点击次数，然后统计出每个种类top3热门的商品。

package cn.spark.study.streaming;

import java.util.ArrayList;
import java.util.List;
import org.apache.spark.SparkConf;
import org.apache.spark.api.java.JavaPairRDD;
import org.apache.spark.api.java.JavaRDD;
import org.apache.spark.api.java.function.Function;
import org.apache.spark.api.java.function.Function2;
import org.apache.spark.api.java.function.PairFunction;
import org.apache.spark.sql.DataFrame;
import org.apache.spark.sql.Row;
import org.apache.spark.sql.RowFactory;
import org.apache.spark.sql.hive.HiveContext;
import org.apache.spark.sql.types.DataTypes;
import org.apache.spark.sql.types.StructField;
import org.apache.spark.sql.types.StructType;
import org.apache.spark.streaming.Durations;
import org.apache.spark.streaming.api.java.JavaPairDStream;
import org.apache.spark.streaming.api.java.JavaReceiverInputDStream;
import org.apache.spark.streaming.api.java.JavaStreamingContext;
import scala.Tuple2;

/**
 * 与Spark SQL整合使用，top3热门商品实时统计
 */
public class Top3HotProduct {

    public static void main(String[] args) {
        SparkConf conf = new SparkConf()
                .setMaster("local[2]")
                .setAppName("Top3HotProduct");  
        JavaStreamingContext jssc = new JavaStreamingContext(conf, Durations.seconds(1));
        
        // 输入日志的格式
        // leo iphone mobile_phone
        
        // 首先，获取输入数据流               
        JavaReceiverInputDStream productClickLogsDStream = jssc.socketTextStream("hadoop1", 9999);
        
        // 然后，应该是做一个映射，将每个种类的每个商品，映射为(category_product, 1)的这种格式
        // 从而在后面可以使用window操作，对窗口中的这种格式的数据，进行reduceByKey操作
        // 从而统计出来，一个窗口中的每个种类的每个商品的，点击次数
        JavaPairDStream categoryProductPairsDStream = productClickLogsDStream
                .mapToPair(new PairFunction() {

                    private static final long serialVersionUID = 1L;

                    @Override
                    public Tuple2 call(String productClickLog)
                            throws Exception {
                        String[] productClickLogSplited = productClickLog.split(" "); 
                        return new Tuple2(productClickLogSplited[2] + "_" + 
                                productClickLogSplited[1], 1);
                    }
                    
                });
        
        // 然后执行window操作
        // 到这里，就可以做到，每隔10秒钟，对最近60秒的数据，执行reduceByKey操作
        // 计算出来这60秒内，每个种类的每个商品的点击次数
        JavaPairDStream categoryProductCountsDStream = 
                categoryProductPairsDStream.reduceByKeyAndWindow(
                        
                        new Function2() {

                            private static final long serialVersionUID = 1L;
                
                            @Override
                            public Integer call(Integer v1, Integer v2) throws Exception {
                                return v1 + v2;
                            }
                            
                        }, Durations.seconds(60), Durations.seconds(10));  
        
        // 然后针对60秒内的每个种类的每个商品的点击次数
        // foreachRDD，在内部，使用Spark SQL执行top3热门商品的统计
        categoryProductCountsDStream.foreachRDD(new Function, Void>() {
            
            private static final long serialVersionUID = 1L;

            @Override
            public Void call(JavaPairRDD categoryProductCountsRDD) throws Exception {
                // 将该RDD，转换为JavaRDD的格式
                JavaRDD categoryProductCountRowRDD = categoryProductCountsRDD.map(
                        
                        new Function, Row>() {

                            private static final long serialVersionUID = 1L;

                            @Override
                            public Row call(Tuple2 categoryProductCount)
                                    throws Exception {
                                String category = categoryProductCount._1.split("_")[0];
                                String product = categoryProductCount._1.split("_")[1];
                                Integer count = categoryProductCount._2;
                                return RowFactory.create(category, product, count);   
                            }
                            
                        });
                
                // 然后，执行DataFrame转换
                List structFields = new ArrayList();
                structFields.add(DataTypes.createStructField("category", DataTypes.StringType, true)); 
                structFields.add(DataTypes.createStructField("product", DataTypes.StringType, true));  
                structFields.add(DataTypes.createStructField("click_count", DataTypes.IntegerType, true));  
                StructType structType = DataTypes.createStructType(structFields);
                
                HiveContext hiveContext = new HiveContext(categoryProductCountsRDD.context());
                
                DataFrame categoryProductCountDF = hiveContext.createDataFrame(
                        categoryProductCountRowRDD, structType);
                
                // 将60秒内的每个种类的每个商品的点击次数的数据，注册为一个临时表
                categoryProductCountDF.registerTempTable("product_click_log");  
                
                // 执行SQL语句，针对临时表，统计出来每个种类下，点击次数排名前3的热门商品
                DataFrame top3ProductDF = hiveContext.sql(
                        "SELECT category,product,click_count "
                        + "FROM ("
                            + "SELECT "
                                + "category,"
                                + "product,"
                                + "click_count,"
                                + "row_number() OVER (PARTITION BY category ORDER BY click_count DESC) rank "
                            + "FROM product_click_log"  
                        + ") tmp "
                        + "WHERE rank<=3");
                

                // 接下来应该将数据保存到redis缓存、或者是mysql db中
                // 然后，配合一个J2EE系统，进行数据的展示和查询、图形报表
                
                top3ProductDF.show();      
                
                return null;
            }
            
        });
        
        jssc.start();
        jssc.awaitTermination();
        jssc.close();
    }   
}

package cn.spark.study.streaming

import org.apache.spark.SparkConf
import org.apache.spark.streaming.StreamingContext
import org.apache.spark.streaming.Seconds
import org.apache.spark.sql.Row
import org.apache.spark.sql.types.StructType
import org.apache.spark.sql.types.StructField
import org.apache.spark.sql.types.StringType
import org.apache.spark.sql.types.IntegerType
import org.apache.spark.sql.hive.HiveContext


object Top3HotProduct {
  
  def main(args: Array[String]): Unit = {
    val conf = new SparkConf()
        .setMaster("local[2]")  
        .setAppName("Top3HotProduct")
    val ssc = new StreamingContext(conf, Seconds(1))
    
    val productClickLogsDStream = ssc.socketTextStream("spark1", 9999)  
    val categoryProductPairsDStream = productClickLogsDStream
        .map { productClickLog => (productClickLog.split(" ")(2) + "_" + productClickLog.split(" ")(1), 1)}
    val categoryProductCountsDStream = categoryProductPairsDStream.reduceByKeyAndWindow(
        (v1: Int, v2: Int) => v1 + v2, 
        Seconds(60), 
        Seconds(10))  
    
    categoryProductCountsDStream.foreachRDD(categoryProductCountsRDD => {
      val categoryProductCountRowRDD = categoryProductCountsRDD.map(tuple => {
        val category = tuple._1.split("_")(0)
        val product = tuple._1.split("_")(1)  
        val count = tuple._2
        Row(category, product, count)  
      })
      
      val structType = StructType(Array(
          StructField("category", StringType, true),
          StructField("product", StringType, true),
          StructField("click_count", IntegerType, true)))
          
      val hiveContext = new HiveContext(categoryProductCountsRDD.context)
      
      val categoryProductCountDF = hiveContext.createDataFrame(categoryProductCountRowRDD, structType)  
      
      categoryProductCountDF.registerTempTable("product_click_log")  
      
      val top3ProductDF = hiveContext.sql(
            "SELECT category,product,click_count "
            + "FROM ("
              + "SELECT "
                + "category,"
                + "product,"
                + "click_count,"
                + "row_number() OVER (PARTITION BY category ORDER BY click_count DESC) rank "
              + "FROM product_click_log"  
            + ") tmp "
            + "WHERE rank<=3")
            
      top3ProductDF.show()
    })
    
    ssc.start()
    ssc.awaitTermination()
  } 
}