Askalon

Grid Application Development and Computing Environment

Performance Monitoring and Analysis

The evolution of parallel and distributed architectures and programming paradigms for performance-oriented program development challenges the state of technology for performance tools. Performance analysis tools should be able to cope with applications for multiple programming models and target architectures. Performance tools must be able to observe performance problems at all levels of a system while relating low-level behavior to the application program. Due to the complexity of the applications and the systems on which the applications are executed, there is a need of collecting, gathering and utilizing monitoring and performance data from many sources, which may be distributed and diverse, in order to understand the performance of the applications. The complexity and quantity of performance measurements are so overwhelming that new performance analysis techniques are required to support efficient, scalable and fast analyses. In addition, performance data needs to be shared and exchanged among different tools. Therefore, techniques and methods to represent and archive performance data and to support tool integration are also importance.

In this topic, our work aims at developing a unified system for performance data integration, instrumentation, measurement, monitoring and analysis for the Grid. The work will be centered on

to support

Moreover, we focus on studying the use of ontology for performance analysis domain. We develop ontology for representing performance data in the Grid with the hope that the proposed ontology will not only serve for data sharing and reuse between performance analysis tools but also increase the automation of performance analysis process.

Contact

Publications

askalon

Overview People Publications Jobs

technology

AGWL Workflow Composition GroudSim

Workflow Execution (Meta) Scheduling Performance Prediction Performance Analysis

Resource Broker Resource Monitoring

Valid HTML 4.0! Valid CSS!