In the thesis, the monitoring and the recording of the consumption of electricity and cooling with the air-conditioning system of the server room of the IT department of the University of Ljubljana is described. Information and communication technology services provided by the University IT (Information Technology) department to the rectorate, university faculties and on these services relying state institutions, such as the Ministry of Education, Science and Sport, are essentially dependent on the reliable electrical supply of ICT (Information and communications technology) equipment in the system room. With the monitoring of the provision of electrical energy, we can predict possible complications that could occur in the future, if we would not take action at appropriate time. Therefore, the quality of the electrical energy supply is essential for the proper operation of the ICT applications.
The stability of electric supply and other important and auxiliary parameters are monitored by the PRTG (Paessler Router Traffic Grapher) monitoring system, which collects and stores the measurements values of thousands of sensors for full insight into the operation.
Thesis presents detailed description of the electricity supply monitoring system. This includes measuring equipment and connections between them. It contains detailed description of the mechanisms and implementation of the Modbus protocol with RS-485 communication, the SNMP (Simple Network Management Protocol) protocol and the PRTG network monitoring system, which is also used for monitoring of other elements of information systems and the communication network at University of Ljubljana.
Thesis also describes methods and importance of data collection and collected measurements of electrical quantities and other parameters, that are critical to ensure a reliable electrical supply and consistent temperature regulation in the server room.
Monitoring of the operation of ICT systems over several years indicates that the data collected can be trusted and that current measured values are accurate. This provides immediate warning in the event of problems and the ability to predict long-term trends and to show daily, weekly and seasonal fluctuations.
The measured results show that it is possible to discover and study more specific features, such as consumption optimization or adverse events prevention.
|