A scalable two-phase top-down specialization approach using Dache in MapReduce framework

摘要

The big-data refer to the extensive disseminated data processing approach that operates on Enormous amount of data. Apache Hadoop is an freely available framework used for big data applications. MapReduce is a programming model. MapReduce framework makes an intermediate data. After finishing the task in MapReduce the completion time of MapReduce jobs is high and data retrieval is low. This paper aims to process large scale data using Dache in a MapReduce framework. In Data aware cache, intermediate results are submitted to the cache manager. Dache improves the completion time of MapReduce jobs, quickly retrieve the data efficiently and enhance the scalability.