Performance-Forecast and Resource-Autonomy Grid Monitoring Architecture (PFRA-GMA)

Grid monitoring is a process to collect information regarding the current and past status of large-scale distributed grid computing resource. It is still a challenge to efficiently cope with the highly complicated issues of monitoring distributed heterogeneous resources of different ownerships in Grids. For that, we propose PFRA-GMA (PerformanceForecast and Resource-Autonomy Grid Monitoring Architecture) which has been developed preliminary in Sceye monitoring system for Scientific Computing Grid (SCGrid) in Chinese Academy of Sciences. It is based on the GMA (Grid Monitoring Architecture) and SOA. The special strength of this implementation comes from the power of its dynamical performance forecast and resource autonomy. Forecast service takes periodic measurements of deliverable resource performance from grid distributed networked resources, and uses numerical models to dynamically generate forecasts of future performance levels. Resource autonomy based on a closed loop control is a good mechanism for efficient monitoring within the scope of same ownership.