Trade-off in desgning Web clusters

被引:24
作者
Menascé, DA [1 ]
机构
[1] George Mason Univ, E Ctr E Business, Fairfax, VA 22030 USA
[2] George Mason Univ, Masters E Commerce Program, Fairfax, VA 22030 USA
关键词
D O I
10.1109/MIC.2002.1036041
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
High-volume Web sites use clusters of Web servers as a way to spread the processing load among various machines, provide increased reliability, reduce response time, and allow for incremental growth in capacity. Some sites have thousands of very inexpensive servers while others use a much smaller number of expensive, high-capacity machines. The described approach provides a simple, but sound, formulation for examining the tradeoffs in this large design space.
引用
收藏
页码:76 / 80
页数:5
相关论文
共 8 条
[1]  
ANDREOLINI M, 2002, ACM SIGMETRICS PERFO, V30
[2]  
ARLITT MF, 1996, P ACM SIGMETRICS 96, P126, DOI DOI 10.1145/233013.233034
[3]   Dynamic load balancing on Web-server systems [J].
Cardellini, V ;
Colajanni, M ;
Yu, PS .
IEEE INTERNET COMPUTING, 1999, 3 (03) :28-+
[4]  
CROVELLA ME, 1996, P ACM SIGMETRICS 96, P160
[5]  
Kleinrock L., 1975, QUEUEING SYSTEMS VOL
[6]   A PROOF FOR THE QUEUING FORMULA - L=LAMBDA-W [J].
LITTLE, JDC .
OPERATIONS RESEARCH, 1961, 9 (03) :383-387
[7]  
Menasce D.A., 2002, CAPACITY PLANNING WE
[8]  
MENASCE DA, 2000, SCALING E BUSINESS T