Matchmaking: Distributed Resource Management for High Throughput Computing

@inproceedings{citeulike:1000729, abstract = {Conventional resource management systems use a system model to describe resources and a centralized scheduler to control their allocation. We argue that this paradigm does not adapt well to distributed systems, particularly those built to support high-throughput computing. Obstacles include heterogeneity of resources, which make uniform allocation algorithms difficult to formulate, and distributed ownership, leading to widely varying allocation policies. Faced with these problems, we developed...}, author = {Raman, Rajesh and Livny, Miron and Solomon, Marvin H. }, booktitle = {HPDC}, citeulike-article-id = {1000729}, keywords = {classification, management, resource}, pages = {140+}, posted-at = {2007-09-28 02:51:57}, priority = {2}, title = {Matchmaking: Distributed Resource Management for High Throughput Computing}, url = {http://citeseer.ist.psu.edu/raman98matchmaking.html}, year = {1998} }

TY - CONF ID - citeulike:1000729 L3 - citeulike-article-id:1000729 TI - Matchmaking: Distributed Resource Management for High Throughput Computing T2 - HPDC SP - 140 N2 - Conventional resource management systems use a system model to describe resources and a centralized scheduler to control their allocation. We argue that this paradigm does not adapt well to distributed systems, particularly those built to support high-throughput computing. Obstacles include heterogeneity of resources, which make uniform allocation algorithms difficult to formulate, and distributed ownership, leading to widely varying allocation policies. Faced with these problems, we developed... KW - classification KW - management KW - resource AU - Raman, Rajesh AU - Livny, Miron AU - Solomon, Marvin PY - 1998/// UR - http://citeseer.ist.psu.edu/raman98matchmaking.html ER -