Spark 如何写入HBase/Redis/MySQL/Kafka

• Partition 是一个可迭代数据集合
  
• Task 本质是作用于Partition的线程
  

object SimpleHBaseClient {
  private val DEFAULT_ZOOKEEPER_QUORUM = "127.0.0.1:2181"

  private lazy val (table, conn) = createConnection

  def bulk(items:Iterator) = {
      items.foreach(conn.put(_))
      conn.flush....
  } 
 ......
}

dstream.foreachRDD{ rdd =>
    rdd.foreachPartition{iter=>
        SimpleHBaseClient.bulk(iter)  
    }
}