spark 方面的的链接

==============SparkStreaming 保证数据准确性

    https://github.com/cpbaranwal/Spark-Streaming-DirectKafka-Examples

    https://github.com/ippontech/spark-kafka-source

    http://aseigneurin.github.io/2016/05/07/spark-kafka-achieving-zero-data-loss.html

    http://blog.youkuaiyun.com/u010454030/article/details/54985740

kafka 10的，有详细的测试记录:  http://blog.youkuaiyun.com/feloxx/article/details/70789000

https://github.com/feloxx/SparkStreaming-DirectKafka010/blob/master/src/main/scala/com/cdp/DealFlowBills2.scala

===========SparkStreaming 自定义receive

http://www.cnblogs.com/ChouYarn/p/7992724.html

=========不错的大数据

https://www.cnblogs.com/smartloli/p/6266453.html