DocumentCode :
2322516
Title :
ELDMA: Enhanced Load Balancing Decision making using Decentralized Mobile Agent framework
Author :
Aramudhan, M. ; Karthikeyan, S. ; Mohan, K. ; Uthaiaraj, Rhymend V.
Author_Institution :
Velammal Eng. Coll., Chennai
fYear :
2008
fDate :
13-15 May 2008
Firstpage :
11
Lastpage :
14
Abstract :
A modified version of LDMA is proposed as an ELDMA (Enhanced Load balancing Decision making using Decentralized Mobile Agent), which distributes the load equally among the Web servers organized in a mesh topology, by a communication media. In LDMA, the rank is assigned to the Web server as and when it enters the cluster. The rank shows the priority of processing a request by the servers in the cluster. Each server involves in the decision making for processing the request by exchanging the message. It causes high communication overhead between the servers. In the proposed scheme, a rank is dynamically assigned to the servers based on CPU processing time and memory utilization. In addition to that the communication overhead is reduced by restricting the message exchange between the servers. The performance of the proposed scheme is evaluated using load distribution, throughput and network traffic.
Keywords :
decision making; file servers; message passing; mobile agents; resource allocation; telecommunication network topology; telecommunication traffic; workstation clusters; ELDMA; Web server cluster; decentralized mobile agent framework; distributed LAN Web server environment; enhanced load balancing decision making; load distribution; memory utilization; mesh topology; message exchange; network throughput; network traffic; Broadcasting; Decision making; Distributed computing; Load management; Local area networks; Mobile agents; Mobile communication; Network servers; Web server; Wide area networks;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer and Communication Engineering, 2008. ICCCE 2008. International Conference on
Conference_Location :
Kuala Lumpur
Print_ISBN :
978-1-4244-1691-2
Electronic_ISBN :
978-1-4244-1692-9
Type :
conf
DOI :
10.1109/ICCCE.2008.4580559
Filename :
4580559
Link To Document :
بازگشت