Spark读取Hbase的数据

val conf = HBaseConfiguration.create()
    conf.addResource(new Path("/opt/cloudera/parcels/CDH-5.4.4-1.cdh5.4.4.p0.4/lib/hbase/conf/hbase-site.xml"))
    conf.addResource(new Path("/opt/cloudera/parcels/CDH-5.4.4-1.cdh5.4.4.p0.4/lib/hadoop/etc/hadoop/core-site.xml"))
    conf.set(TableInputFormat.INPUT_TABLE, "FLOW")

    //添加过滤条件，年龄大于 18 岁
    //val scan = new Scan()
    //conf.set(TableInputFormat.SCAN, convertScanToString(scan))
    /*
    scan.setFilter(new SingleColumnValueFilter("basic".getBytes, "age".getBytes,
      CompareOp.GREATER_OR_EQUAL, Bytes.toBytes(18)))
    */

    val usersRDD = sc.newAPIHadoopRDD(conf, classOf[TableInputFormat],
      classOf[org.apache.hadoop.hbase.io.ImmutableBytesWritable],
      classOf[org.apache.hadoop.hbase.client.Result])

    val data1 = usersRDD.count()

    val sf = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss.SSSSS")

    println("data length:" + data1)

    var map = HashMap[String, HashMap[String, collection.mutable.ArrayBuffer[Double]]]()

    usersRDD.collect().map {
      case (_, result) =>
        val key = Bytes.toInt(result.getRow)
        println("Key:" + key)
        val ip = Bytes.toString(result.getValue("F".getBytes, "SADDR".getBytes))
        val port = Bytes.toString(result.getValue("F".getBytes, "SPORT".getBytes))
        val startTimeLong = Bytes.toString(result.getValue("F".getBytes, "STIME".getBytes))
        val endTimeLong = Bytes.toString(result.getValue("F".getBytes, "LTIME".getBytes))
        val protocol = Bytes.toString(result.getValue("F".getBytes, "PROTO".getBytes))
        val sumTime = Bytes.toString(result.getValue("F".getBytes, "DUR".getBytes))
        val sum = Bytes.toString(result.getValue("F".getBytes, "DBYTES".getBytes)).toDouble

        println("ip:" + ip + ",port:" + port + ",startTime:" + startTimeLong + ",endTime:" + endTimeLong + ",protocol:" + protocol + ",sum:" + sum)

        //ip+port+udp，14:02 14:07 List
        //ip+port+tcp，15:02 15:07 List
        val startTimeDate = sf.parse(startTimeLong)
        val endTimeLongDate = sf.parse(endTimeLong)
        val startHours = startTimeDate.getHours
        val startMinutes = startTimeDate.getMinutes

        val endHours = endTimeLongDate.getHours
        val endMinutes = endTimeLongDate.getMinutes

        val key1 = ip + "_" + port + "_" + protocol
        println("key1:" + key1)

        val key2 = startHours + ":" + startMinutes + "_" + endHours + ":" + endMinutes

        println("key2:" + key2)

        val tmpMap = map.get(key1)

        if (!tmpMap.isEmpty) {
          println("--------------------map is not null:" + tmpMap.size + "--------------------")
          val sumArray = tmpMap.get.get(key2)
          if (!sumArray.isEmpty) {
            sumArray.get += sum
          }
        } else {
          println("--------------------map is null--------------------")
          //如果当前Key不存在的话，是一个全新的Ip
          val sumArray = collection.mutable.ArrayBuffer[Double]()
          sumArray += sum

          val secondMap = HashMap[String, collection.mutable.ArrayBuffer[Double]]()
          secondMap += (key2 -> sumArray)
          map += (key1 -> secondMap)
        }
        map
        println("map size-----------------:" + map.size)
    }

    println("map size:" + map.size)

    map.map(e => {
      println("--------------------Statistics start --------------------")
      val resultKey1 = e._1
      val resultVal1 = e._2
      println("resultKey1:" + resultKey1)
      resultVal1.foreach(f => {
        val resultKey2 = f._1
        val resultVal2 = f._2
        println("resultKey2:" + resultKey2)
        println("-----------------resultVal2:" + resultVal2.length)

        resultVal2.map(f=>{
            println("------------------------f:"+f)
        })

        val dataArray = resultVal2.map(f => Vectors.dense(f))

        val summary: MultivariateStatisticalSummary = Statistics.colStats(sc.parallelize(dataArray))

        //
        println("--------------------mean:" + summary.mean + " --------------------")
        println("--------------------variance:" + summary.variance + " --------------------")

        println("--------------------mean apply 0:" + summary.mean.toArray.apply(0) + " --------------------")
        println("--------------------variance apply 0:" + summary.variance.apply(0) + " --------------------")

        val upbase = summary.mean.toArray.apply(0) + 1.960 * Math.sqrt(summary.variance.apply(0))
        val downbase = summary.mean.toArray.apply(0) - 1.960 * Math.sqrt(summary.variance.apply(0))
        println("------------------- " + upbase + " ---------- " + downbase)
        val df = new DecimalFormat(".##")
        val upbaseString = df.format(upbase)
        val downbaseString = df.format(downbase)
        //resultMap.put(key, value)
        val result3 = HashMap[Double, Double]()
        //result3 +=(upbase -> downbase)
        println("ip port:" + resultKey1 + ",time:" + resultKey2 + ",upbase:" + upbase + ",downbase:" + downbase)
      })
    })

    println("--------------------baseLine end --------------------")
    sc.stop()