EQUILOAD: a load balancing policy for clustered web servers

被引:54
作者
Ciardo, G [1 ]
Riska, A [1 ]
Smirni, E [1 ]
机构
[1] Coll William & Mary, Dept Comp Sci, Williamsburg, VA 23187 USA
基金
美国国家科学基金会; 美国国家航空航天局;
关键词
web server; workload allocation; queue policy; processing time;
D O I
10.1016/S0166-5316(01)00049-9
中图分类号
TP3 [计算技术、计算机技术];
学科分类号
0812 ;
摘要
We present a new strategy for the allocation of requests in clustered web servers, based on the size distribution of the requested documents. This strategy, EQUILOAD, manages to achieve a balanced load to each of the back-end servers, and its parameters are obtained from the analysis of a trace's past data. To study its performance, we use phase-type distribution fittings and solve the resulting models using a new solution method for M/PH/1 queues that only requires solution of linear systems. The results show that EQUILOAD greatly outperforms random allocation, performs comparably or better than the Shortest Remaining Processing Time and Join Shortest Queue policies and maximizes cache hits at the back-end servers, therefore behaving similarly to a "locality-aware" allocation policy, but at a very low implementation cost. (C) 2001 Elsevier Science B.V. All rights reserved.
引用
收藏
页码:101 / 124
页数:24
相关论文
共 30 条
[1]  
Almeida V, 1996, PROCEEDINGS OF THE FOURTH INTERNATIONAL CONFERENCE ON PARALLEL AND DISTRIBUTED INFORMATION SYSTEMS, P92, DOI 10.1109/PDIS.1996.568672
[2]  
Arlitt M., 1999, WORKLOAD CHARACTERIZ
[3]  
ARLITT MF, 1996, P ACM SIGMETRICS 96, P126, DOI DOI 10.1145/233013.233034
[4]  
Barford P., 1998, Performance Evaluation Review, V26, P151, DOI 10.1145/277858.277897
[5]   Changes in Web client access patterns: Characteristics and caching implications [J].
Barford P. ;
Bestavros A. ;
Bradley A. ;
Crovella M. .
World Wide Web, 1999, 2 (1-2) :15-28
[6]  
Cao P., 1997, P USENIX S INT TECHN
[7]   Characterizing temporal locality and its impact on web server performance [J].
Cherkasova, L ;
Ciardo, G .
NINTH INTERNATIONAL CONFERENCE ON COMPUTER COMMUNICATIONS AND NETWORKS, PROCEEDINGS, 2000, :434-441
[8]   ETAQA: an efficient technique for the analysis of QBD-processes by aggregation [J].
Ciardo, G ;
Smirni, E .
PERFORMANCE EVALUATION, 1999, 36-7 :71-93
[9]  
CIARDO G, 1999, NUMERICAL SOLUTION M, P21
[10]   Analysis of task assignment policies in scalable distributed web-server systems [J].
Colajanni, M ;
Yu, PS ;
Dias, DM .
IEEE TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS, 1998, 9 (06) :585-600