Tackling Heterogeneous Traffic in Multi-access Systems via Erasure Coded Servers

Abstract

In cloud systems, the number of servers is fixed, and each server is usually dedicated to a job type. However, the traffic of various jobs can vary across time, and hence, a rigid allocation of servers does not perform well. In this paper, we propose the addition of erasure-coded servers that can flexibly serve multiple job types without additional storage cost. We show that coding expands the service capacity region, thus enabling the system to handle variable traffic. Moreover, we characterize the response time of the coded system in various arrival rate regimes and show that adding even a small number of coded servers can significantly reduce the mean response time.