Multiresource Allocation: Fairness–Efficiency Tradeoffs in a Unifying Framework

Quantifying the notion of fairness is underexplored when there are multiple types of resources and users request different ratios of the different resources. A typical example is data centers processing jobs with heterogeneous resource requirements on CPU, memory, network bandwidth, etc. In such cases, a tradeoff arises between equitability, or “fairness,” and efficiency. This paper develops a unifying framework addressing the fairness-efficiency tradeoff in light of multiple types of resources. We develop two families of fairness functions that provide different tradeoffs, characterize the effect of user requests´ heterogeneity, and prove conditions under which these fairness measures satisfy the Pareto efficiency, sharing incentive, and envy-free properties. Intuitions behind the analysis are explained in two visualizations of multiresource allocation. We also investigate people´s fairness perceptions through an online survey of allocation preferences.