A QoS-aware resource allocation framework in virtualised cloud environments

In cloud platforms, resource allocation service plays an important role for running user applications efficiently. However, current allocation mechanism in many cloud platforms only provides best-effort service for user’s jobs, which means that user’s quality-of-service (QoS) requirements can not be well guaranteed. In this paper, we present a novel QoS-aware resource allocation framework, which applies feedback control technique to achieve the goal of fair resource allocation between multiple virtual machine (VM) instances. Theoretical analysis of the proposed control model has proven that it can meet the feasibility and stability requirements. Experimental results conducted on a real-world cloud platform show that the proposed resource allocation framework can significantly improve the effective resource utilisation as well as the overall task execution efficiency. In addition, the proposed framework also shows better robustness when the cloud platform is in presence of highly dynamic workload. Copyright © 2019 Inderscience Enterprises Ltd.