Improve Constrained Recourse Fairness in Cloud Computing with Bottleneck-Aware Allocation

Abstract

Listen

Translate article

Fair resource is a key building block of any shared computing system. Recently fair division theory has emerged as a promising approach for the allocation of multiple computational resources among agents. Recent research has discussed efficiency and fairness requirements and identified a number if desirable scheduling objectives including so-called dominant resource fairness(DRF). DRF is not good between fair and resource utilization. A new allocation model Balancing Fairness and Efficiency with BottleneckAware Allocation(BAA) find good appropriate balance between fairness to the clients and maximizing system utilization. But BAA without taking into account the weight value and maximum number of tasks which are users need to run. We propose the IBAA fair allocation mechanism. IBAA has lots of good properties, it satisfies DSI, PE and EF. We construct IBAA mechanisms that provably satisfy properties, and analyze the performance. We believe that our work informs the design of superior multiusers system, and at the same time expands the scope of fair division theory by initiating the study of dynamic and fair resource allocation mechanisms.