Abstract
Today, datacenter providers spend a lot of money on inter-datacenter WANs to transmit traffic flows between geographically distributed datacenter sites. It is very important for datacenter providers to design resource scheduling schemes to schedule inter-datacenter networking resources to satisfy data transfer requests and achieve maximum performance or economic benefits. In this article, we conduct a review study on research efforts being conducted on the inter-datacenter networking resource scheduling problem. Our survey study is presented in three parts. The first part focuses on formulating the problem theoretically, including the definition and features of a data transfer, network models of inter-DC WANs, scheduling objectives, and scheduling dimensions. We also summarize a conceptual implementation architecture of scheduling systems. Second, we classify existing schemes according to their objectives and scheduling dimensions and then explain how they formulate and solve their own problems. In the third part, we examine practical challenges in developing scheduling systems and also point out some future directions. Since we do not see any survey focusing on the scheduling problem of inter-DC WANs, we believe this article can provide some help for research developments in this important field by organizing logically the recent research efforts from industry and academia.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call