The field of network management is of strategic importance to modern computer networks. People are becoming increasingly reliant on computer networks for everyday tasks such as personal and corporate communication, banking, commerce, and shopping. These services are expected to be available at all times, and the cost of failing to meet this expectation is often substantial.
Unfortunately, network management is a field that is fraught with intricacies and problems. One of the most significant of these arises as a result of the way networks have developed over time; the problem of running a hetrogeneous network utilising components from many different manufacturers has frustrated many attempts at providing a unified way of managing and monitoring networks.
Most current network monitoring and management tools suffer from a number of common limitations, and it is these limitations that form the focus of this work.
This project sets out to identify several common problem areas that are not adequately addressed by current network monitoring techniques. It uses the network at Rhodes University as an example of a typical campus network, and notes that it suffers from several shortfalls in terms of its current network monitoring strategy. These shortfalls represent problems that are typical of many other organisations, and as a result, the University forms a useful platform on which to investigate these areas.
Having identified four such problems at the University, this project aims to find novel solutions that provide the University's network administrators with the means to address these issues.
These solutions are then investigated in light of the Open Systems Interconnect network management model to determine how they fit together and form a more complete, and more universally applicable network monitoring solution.