|
Title:
|
REPLICA MANAGEMENT IN CONTENT DISTRIBUTION NETWORKS |
|
Author(s):
|
Vibha Tripathi , Yuri Breitbart |
|
ISBN:
|
ISSN: 1645-7641 |
|
Editors:
|
Pedro Isaías |
|
Year:
|
2006 |
|
Edition:
|
V IV, 1 |
|
Keywords:
|
Primary server, Candidate server, Content Replication, Replica(s), Response Time, Consistency |
|
Type:
|
Journal Paper |
|
First Page:
|
43 |
|
Last Page:
|
54 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
With the rapid increase in the number of Internet users, fast accessibility of web documents has become an important issue. Web Replication is a well-known approach to reduce client latencies and increase the availability of a website. Various Content Distribution Networks (CDNs) are in the market nowadays to facilitate replication of content for web content providers, e.g. Akamai, Digital Island. These systems have been very successful and have made the problem of replication a more important problem for research. The replication algorithms these Content Distribution Networks use, differ in the manner they make decisions as to when, where and what content to replicate in order to maximize the availability of a certain content. The decision parameters used in these algorithms play an important role in the efficiency of the system. We studied several such algorithms to determine which parameters are statistically more significant than the others and we propose a new algorithm for web content replication. We show using simulations that our algorithm performs better than the other algorithms. We also propose a new architectural framework for Content Delivery Networks that would maximize performance of servers that use our replication algorithm. |
|
|
|
|
|
|
