This paper addresses the problem of a unified framework for resource scheduling in shared heterogeneous cluster computing systems. Most research in job scheduling has focused solely on the allocation of processors to jobs. However, since I/O is also a critical resource for many parallel jobs, the allocation of processor and I/O resources must be coordinated to allow the system to operate most effectively. To this end, we propose a vertically distributed framework that defines bindings between computing power and data resources. Based on this framework, we propose two approaches to scheduling CPU-I/O bound jobs and study their performance under various system and workload parameters.
History
Pagination
43-48
Location
Reno, Nev.
Start date
2003-08-13
End date
2003-08-15
ISBN-10
1-880843-48-X
Language
eng
Publication classification
EN.1 Other conference paper
Editor/Contributor(s)
Yoo S-M, Youn HY
Title of proceedings
PDCS 2003 : Proceedings of the 16th International Conference on Parallel and Distributed Computing Systems
Event
PDCS Parallel and Distributed Computing Systems. International Conference (16th : 2003 : Reno, Nevada)
Publisher
International Society for Computers and the Applications (ISCA)