Klemm M, Veldema R, Philippsen M (2008)
Publication Language: English
Publication Type: Conference contribution, Original article
Publication year: 2008
Edited Volumes: Proceedings of the IASTED International Conference on Parallel and Distributed Computing and Systems
City/Town: Anaheim, CA, USA
Pages Range: 219-224
Conference Proceedings Title: Proceedings of the 20th IASTED International Conference on Parallel and Distributed Computing and Systems (PDCS'08)
Event location: Orlando, FL, USA
ISBN: 978-0-88986-773-4
URI: http://www2.informatik.uni-erlangen.de/publication/download/OGRE-PDCS.pdf
Grid computing promises to seamlessly deliver computing power to computational scientists. However, scientists are faced with the problem of finding computing resources for their application and they have to cope with the systems' batch operation mode that causes premature termination of applications if they exceed the reserved resource limits. In this paper, we propose a middleware layer (OGRE) for a seamless Grid environment. OGRE automates the process of resource discovery and performs resource selection. Resource selection uses a cost-based migration strategy that allocates free computing resources on accessible clusters, i. e., OGRE automatically checkpoints, transfers the snapshot from one cluster to the next, and restarts the application there. A Grid-wide file system handles data accesses of migrated applications. OGRE can easily be adapted to new requirements due to its component-based architecture.
APA:
Klemm, M., Veldema, R., & Philippsen, M. (2008). An Automatic Cost-based Framework for Seamless Application Migration in Grid Environments. In Proceedings of the 20th IASTED International Conference on Parallel and Distributed Computing and Systems (PDCS'08) (pp. 219-224). Orlando, FL, USA, US: Anaheim, CA, USA.
MLA:
Klemm, Michael, Ronald Veldema, and Michael Philippsen. "An Automatic Cost-based Framework for Seamless Application Migration in Grid Environments." Proceedings of the 20th IASTED International Conference on Parallel and Distributed Computing and Systems (PDCS'08), Orlando, FL, USA Anaheim, CA, USA, 2008. 219-224.
BibTeX: Download