摘要
结合电信增值业务领域中对大数据处理的实际需求,对现有主流的分布式大数据处理架构(Hive、Impala、Spark)的核心进行分析与实测,比较它们在大数据处理过程中的优劣及适用的场景,从而为大数据分析所面临的架构适用性选型提供参考。
A comparison of three open source distributed computing frameworks for big data （Hive, Impala and Spark） was conducted. Tests were run to evaluate the performance aiming at real business demands. The cost of implementation to meet business requirements was also discussed.
出处
《电信科学》
北大核心
2015年第7期152-157,共6页
Telecommunications Science
