International Journal of Advanced Research in Computer Science and Software Engineering (IJARCSSE)
In a high performance cluster environment, usage of an agent based monitoring will consume CPU cycle and memory footprint. Currently host OS monitoring and hardware monitoring are done separately. Hardware monitoring can be done using the IPMI stack. Host OS monitoring can be done using Nagios, Ganglia or a similar software. In order to develop an agent free monitoring tool for HPC cluster monitoring, the authors integrated hardware monitoring tool IPMI (Intelligent Platform Management Interface) stack with Nagios. The integrated framework abstracts the hardware monitoring of the server to the Nagios GUI web interface.