Healthcare Data Analysis using Dynamic Slot Allocation in Hadoop

Abstract

In this new era of big data even health care needs to be modernized, this includes that the health care data should be properly analyzed so that we can deduce that in which group or gender, diseases attack the most. This gigantic size of analytics will need large computation which can be done with help of distributed processing, Hadoop. MapReduce, a popular computing paradigm for large-scale data processing in cloud computing. However, the slot-base MapReduce system (e.g., Hadoop MRv1) due to its unoptimized resource allocation, can suffer from poor performance. To address it, the framework in this paper optimizes the resource allocation. Due to the static pre-configuration of distinct map slots and reduce slots which are not fungible, many a times slots can be severely under-utilized. This is because map slots might be fully utilized while reduce slots may remain empty, and vice-versa. We propose an alternative technique called Dynamic Hadoop Slot Allocation by keeping the slot-based allocation model. It relaxes the slot allocation constraint and allows slots to be reallocated to either map or reduce tasks depending on their needs. The framework's use will provide multipurpose beneficial outputs which include: getting the health care analysis in various forms. Thus this concept of analytics should be implemented with a view of future use.