원문정보
초록
영어
Virtualization-based cloud computing platforms allow multiple virtual machines (VMs) running on the same physical machine. Efficient allocation of limited underlying resources has been a key issue. In order to improve the CPU resources utilization, this paper presents a workload-aware CPU resources scheduling method (WARS). WARS uses the allocated credits and consumed credits to diagnose the CPU resources requirements of VMs and dynamically adjusts CPU resources according to the requirements of VMs. The adjustment of CPU resources is converted into increased or decreased weights of VMs. The implementation of WARS is confined to the VMM layer, without VM dependency. Our evaluation shows that WARS can improve the overall utilization of CPU resources.
목차
1. Introduction
2. Background
3. The WARS Scheduling Model
3.1. WARS Architecture
3.2. Resources Requirement Diagnose Model
3.3. Resource Scheduling Model
3.4 The Evaluating Model
4. Performance Evaluation
4.1. Experiment Setup
4.2. Performance Evaluation
5. Related Work
6. Conclusions
Acknowledgements
References