The goal of this work is to research various approaches towards forecasting / trending of future performance in IPv4/IPv6 based networks based on the past data to help the network administrators to identify potential problems before they occur. Propose a set of data to be collected by using protocols currently supported by network devices and statistical methods for their correlation and prediction of future trends. Part of the work is implementation of a prototype application that will demonstrate a validity of the results in a typical real-life scenario. It assumes knowledge of networking and network protocols as well as the ability to program in C/C++ and/or C#/.NET technologies. Third party libraries may be used to avoid duplication of functionality and third party tools are expected to be used to gather data about the traffic in the network. The work will be consulted and co-led with SolarWinds, an industrial partner of Faculty of Informatics. It can be written also in Czech language with executive summary in English.