Hardware and software surveillance
MetadataVis full innførsel
The surveillance of computer equipment is an important task at the European Laboratory for Particle Physics (CERN). Given the nature and size of its installations, constant feedback about the health state of its hosts is vitally important to ensure the seamless operation of particle accelerators and other facilities. Historical data of errors is needed to understanding and debug these errors. This data can be used to prevent future similar errors. CERN are using a wide rage of Operating Systems running of many different types of hardware. To survey these equipment CERN need a well-tuned and scaleable applications. This thesis describes the current surveillance system (XCLUC), a new surveillance system (LEMON) and how these can be used together. This thesis also describes some important add-ons to the both surveillance system for current and future use of the surveillance systems. XCLUC is designed to use the minimal of system resources when testing the current state of the computer and when sending the results of this tests to a central server. LEMON is more a general module-based system; the system does not take into account how much of the system resources it uses. This thesis concludes that XCLUC is the best surveillance system for all the small computer equipment, like disk-less PowerPC s. LEMON has the best historical view and is the best system to also analyze previous errors. A combination of XCLUC and LEMON would make the best surveillance system of the computer equipment at CERN.