Abstract
Businesses and individuals often suffer from significant amount of damage as a result of network failures, and that is the reason why network fault tolerant mechanism is an indispensable condition for network design and management. The most common failure is on the network equipments itself. Network equipment is composed of hardware and software parts. There are lots of possible failure points within both parts and expectation to the time of failure is generally difficult. Moreover, network equipments located at the entrance of a network play an important role in the availability and reliability of the network inside that entrance. Therefore, the thesis design and implement a fault-tolerant system in accordance with the network equipments in the entrance of a network. The purposed system architecture could provide fault-tolerance for multiple network equipments simultaneously and has the ability to switch to redundant equipment very fast without management overhead. On the situation that no redundant device exists, the fault-tolerant system could bypass the forwarding path to survive the network connections. The thesis adopts the Intel IXDP1200 Network Processor [1] as development platform to implement the proposed system. Besides, we use hardware testing to demonstrate the performance of purposed fault-tolerant system. Finally, the thesis mentions a few applications that could be implemented under the system architecture as future works.