With the vigorous development of network infrastructure, ISPs provide wider network bandwidth and lower price for users to lease the network. Even though in pursuit of wide-bandwidth service, it does not mean to have better quality as well. In order to consider the economic benefits and to provide uninterrupted services, the so-called redundancy mechanism is derived. This mechanism integrates multiple sets of servers and multiple lines to achieve the purpose of uninterrupted services.
The redundancy mechanism is divided into hot-standby-redundancy and cold-standby-redundancy, where the load balancing system is an important application of hot-standby-redundancy. Generally, the host redundancy uses a single high efficiency servers or cluster servers to achieve uninterrupted services, while the line redundancy uses a number of lines provided by different ISPs to spread the traffic load using various load balancing methods and keep the network free from interruption.
In different network environments and services, there are many algorithms for the load-balancing technology and traffic allocation, and each has its own advantages and disadvantages. This thesis takes the distributed network architecture for example, which is commonly applied in a government organization, and explores the pros and cons of various redundancy mechanisms. In addition, this thesis develops a redundancy mechanism which can improve network availability, strengthen information security management, and be implemented with a lower budget to achieve load balancing purpose.
