Resource Monitoring in ASKALON
Resource monitoring supports the monitoring of Grid resources by integrating and extending existing Grid resource monitoring tools and by developing new techniques for Grid resource monitoring (e.g. rule-based monitoring). To support resource monitoring, we exploit the sensor networks model to provide and integrate various types of performance and monitoring data in a unified system. Regardless of the monitoring data of Grid infrastructures or applications, data is treated and accessed in the same way. Both event-driven and demand-driven sensors are developed to serve different purposes. Sensors also use a rich set of rules and inference engine to control the monitoring. In order to balance tradeoffs between interoperability and performance, we employ both Grid service-based operations and TCP-based stream data delivery. Self-managing and P2P features are exploited to increase the scalability and fault-tolerance, and to foster distributed, group-based and automatic data query and subscription of performance and monitoring data.
This work will be centered around a self-managing sensor-based middleware for Grid monitoring and performance data integration within the SCALEA-G system .
Further information about our research on performance monitoring and analysis themes.