This paper presents the overall design of a multi-agent framework for tuning the performance of an application executing in a distributed environment. The multi-agent framework provides services like resource brokering, analyzing performance monitoring data, local tuning and also rescheduling in case of any performance problem on a specific resource provider. The paper also briefly describes the implementation of some part of the framework. In particular, job migration on the basis of performance monitoring data is particularly highlighted in this paper.
@article{arxiv.1005.2027,
title = {A Multi-agent Framework for Performance Tuning in Distributed Environment},
author = {Sarbani Roy and Saikat Halder and Nandini Mukherjee},
journal= {arXiv preprint arXiv:1005.2027},
year = {2010}
}
Comments
International Conference On High Performance Computing HiPC 2005, Posters Web Proceedings, Goa, India, December 18-21, 2005