Imagine if you would like to deploy your new simulation online and your systems can just take care of itself. Needed web interface could be generated, database schema objects could be created, simulation programs could be installed and configured on your execution servers, data analysis and reports could be generated automatically. In this paper, we describe the design and implementation of a self-manageable multi-tier infrastructure to support scientific simulations. This infrastructure demonstrates not only the successful integration of web servers, execution servers, database servers, reports servers, data warehousing and data mining, but also the ability to achieve self manageability, including self-configuring, self-healing, self-protecting and self-optimizing. A scientific simulation program, NOM, is employed to demonstrate the effectiveness of this infrastructure.
Yingping Huang, Xiaorong Xiang, Gregory R. Madey