本文介绍如何在E-MapReduce的Hadoop集群运行Spark作业、统计单词个数并将结果写入MySQL。Spark Streaming作业写入MySQL的方法与此类似。

Spark接入MySQL

示例代码如下。
val input = getSparkContext.textFile(inputPath, numPartitions)
    input.flatMap(_.split(" ")).map(x => (x, 1)).reduceByKey(_ + _)
      .mapPartitions(e => {
        var conn: Connection = null
        var ps: PreparedStatement = null
        val sql = s"insert into $tbName(word, count) values (?, ?)"
        try {
          conn = DriverManager.getConnection(s"jdbc:mysql://$dbUrl:$dbPort/$dbName", dbUser, dbPwd)
          ps = conn.prepareStatement(sql)
          e.foreach(pair => {
            ps.setString(1, pair._1)
            ps.setLong(2, pair._2)
            ps.executeUpdate()
          })

          ps.close()
          conn.close()
        } catch {
          case e: Exception => e.printStackTrace()
        } finally {
          if (ps != null) {
            ps.close()
          }
          if (conn != null) {
            conn.close()
          }
        }
      Iterator.empty
    }).count()

附录

完整示例代码，请参见Spark写入MySQL。