SCahS: Synchronized Caching Strategies for Incremental Computations on Hadoop

Data for various applications are evolving continuously either online or offline way. The business communities extract this available big data and search the potential business opportunities using it. For sophistication of this huge data many data processing systems are designed by software companies; but neither fulfills efficient updates or enhance efficiency as expected by business community?s demand. These systems lack in performance and do not support compatibility with existing distributed computing environments like Hadoop MapReduce. In addition to these drawbacks it requires to develop only application specific algorithms which need to be replaced dynamically by programmers resulting creating complexity in design and code. The solution is planned to minimize computational time, wastage of input and output bandwidth and optimize resources namely central processing unit, memory utilization for incremental data. Another aspect is also taken care is that the design of proposed system will be simple, easy to understand and execute and fulfill the expectations demanded by users. The system saves intermediately generated data for a specific application by caching it. The result obtained from the proposed system reduces significant computational time and saves memory space.