A web cluster simulator for performance analysis of the ALBM cluster system

  • Authors:
  • Eunmi Choi;Dugki Min

  • Affiliations:
  • School of Business IT, Kookmin University, Seoul, Korea;School of Computer Science and Engineering, Konkuk University, Seoul, Korea

  • Venue:
  • AIS'04 Proceedings of the 13th international conference on AI, Simulation, and Planning in High Autonomy Systems
  • Year:
  • 2004

Quantified Score

Hi-index 0.00

Visualization

Abstract

A distributed server cluster system is a cost-effective solution to provide scalable and reliable Internet services. In order to achieve high qualities in service, it is necessary to tune the system varying configurable parameters and employed algorithms that significantly affect system performance. In this purpose, we develop a simulator for performance analysis of a traffic-distribution cluster system, called the ALBM (Adaptive Load Balancing and Management) cluster. In this paper, we introduce the architecture of the proposed simulator. Major design considerations are given to the flexible structures that can be easily expanded for adding new features, such as new workloads and scheduling algorithms. With this simulator, we perform two simple cases of performance analysis: one is to find appropriate overload and underload thresholds and the other is to find the suitable scheduling algorithm for a given workload.