A Markov-Chain Model Based Study of Distributed Weighted Round-Robin Scheduling for Data Centers

Abstract

Data centers spread around the world at different geographical location with the variation of time and location. This paper offers distributed weighted round robin (DWRR) scheduling algorithm for large-scale distributed data centers using Markov Chain Model. DWRR provides a platform to achieve fairness for all data centers. This proposed algorithm evaluates and optimizes the performance that reduces the operation costs, balances the load effectively, improves fairness and produces maximum throughput of data centers to study the transition behaviour of threads among different data centers.