Font Size: a A A

Research On High Availability Load Balancing Algorithms Of WEB Sites Under Multi-entry

Posted on:2013-02-12Degree:MasterType:Thesis
Country:ChinaCandidate:D WuFull Text:PDF
GTID:2218330374464204Subject:Computer system architecture
Abstract/Summary:PDF Full Text Request
Nowadays, the amount of informations and websites's traffic have been increasing in a geometrical progression, congestion and overload of services have been challenging network bandwidth and processing capabilities of web server. Network cluster is a solution to high-load websites. Multiple web servers are combined in certain connections, and provide services efficiently and reliably. Load balancing technology mainly dwells on how to connect servers and balance traffic distributions. In general, architecture of load balancing depends on specific requirements and constraints of business, and should be applied different strategies in different occasions. Existing exchanging problems among China's Internet service providers also spawn development of multi-link load-balancing technology. The purpose of the article is to ensure efficiency and availability of multi-portal web services by improving balancing strategies.Starting with high availability of Web sites under multi-entry, and by executing load balancing strategies at different levels, the author assumes creative works as following:proposing improved strategy of round-robin scheduling and Trunk; designing and implementing an modified gathering scheme of network resources and routing based on ant colony;putting forwarding IP intelligent recognition technology for multi-link load-balancing.Tests expatiate that the solutions above can expand bandwidth, raise utilization rate of resources, and schedule and balance websites's traffic. By them websevices's efficiency and availability can be gradually enhanced. The solutions should be a good choice for many large and medium-sized websites.
Keywords/Search Tags:website load balancing, availability, Trunk, ant colony, SNMP
PDF Full Text Request
Related items