Bahram Aryana - Ph.D. Candidate at Facultyof Computer Engineering,Islamic Azad UniversityCentral Tehran Branch
Behnaz Nahvi - AssociateProfessor at Faculty ofComputer Engineering,Islamic Azad UniversityKaraj Branch
Erfane Nowruzi - Faculty of ComputerEngineering, Islamic AzadUniversity Qeshm Branch

This study investigates the effects of theYARN on the HaOLAP Hadoop and tries to assesshow different types of schedulers of the YARN canbe used to achieve the optimum result to improvethe performance of the HaOLAP Hadoop. In thefirst step, HaOLAP Hadoop v.۱.۰ is compared withHaOLAP Hadoop v.۳.۲.۱ augmented with YARN,and all three major schedulers namely FIFO, Fair,and Capacity are separately used and their effectson execution time are evaluated. Accordingly, threecollections of data called 𝑪𝟏, 𝑪𝟐, and 𝑪𝟑 containing𝟏𝟎𝟓, 𝟏𝟎𝟔, and 𝟏𝟎𝟕 items respectively from medicalclinic records are selected and evaluated underthree operations 𝑨𝟏, 𝑨𝟐, and 𝑨𝟑. In the second stepperformance of the schedulers is compared andtheir cons and pros are assessed. Finally, somesuggestions to achieve the optimum results toaugment HaOLAP Hadoop with YARN arepresented based on this study’s results.

YARN, HaOLAP, Hadoop, Schedulers,FIFO, Fair, Capacity, MapReduce