Hugepage Swappiness functions for optimization of the search graph algorithm using Hadoop framework

摘要

21st century is best known as technology centaury and the advancement in technology has helped in close knit networking of people in the world. With the advent of internet & social networking sites, connecting with people is a click away. These advancements warranted deep research in social networking [1] and its related technologies. The social networking sites store humongous data and when a user searches somebody, a huge processing happens in the background, this search operation has to be swift to enhance the user's experience. Current research is focused on the optimization of the search operation on Hadoop framework [3], two memory functions "Hugepage" & "Swappiness" of the system are selected for optimization of search algorithm. The search operation is optimized by processing the BigData of size 2GB to 10GB and varying the Swappiness (behave as a virtual memory) of the system in phase-1 and by varying the Hugepage between 10–100% and selecting the optimal points of Swappiness (results of pahse-1) in phase-2.