Passive Network Measurements For Troubleshooting And Testing

With the fast growing of networks and massive progress of the technology around the world, security and the need to ensure that data transfer in the networks were done efficiently remains indispensable. Many companies and operators have transferred important information across networks but no data leakage is tolerated. This thesis discusses the problem of identifying and solving holes in a particular network by using passive network measurements.It is found that the fundamental measuring techniques of passive monitoring have remained unchanged, only the link speeds and technologies have renewed it. With modern home computing comes the possibility of doing end-to-end measurements easily. There are new amended methods to measure things easier, more reliably, and requiring smaller amounts of captured data. Troubleshooting the networks does not need to be done just by searching reasons for problems randomly.In the "small enterprise network" the exclusion model was used partly for searching reasons for problems. It is also good to remember that in troubleshooting the use of a specific particular model should not be the self purpose but the most important thing is to solve the reason for the problem. But it has to be notified that there is not only one proper model or a combination of models which would suit for every case. In addition, if the problem domain is small, it could be wise to search the reason with the help of intuition and not to use any model at all. The value of these models grows when the problem domain and cases enlarge and become more complicated. The largest problem still remains in time-related multipoint measurements:the availability of cost effective accurate clock synchronization methods for PCs. In addition, capturing data fully from the link speed up to 1Gbit/s, is still a problem. Passive packet monitoring is a powerful way for troubleshooting and testing network devices.