Please use this identifier to cite or link to this item:

Dynamic fractional resource scheduling for cluster platforms

File Description SizeFormat 
Stillwell_Mark_r.pdfVersion for non-UH users. Copying/Printing is not permitted653.92 kBAdobe PDFView/Open
Stillwell_Mark_uh.pdfVersion for UH users620.48 kBAdobe PDFView/Open

Item Summary

Title: Dynamic fractional resource scheduling for cluster platforms
Authors: Stillwell, Mark Lee
Keywords: job scheduling
homogeneous computational clusters
Issue Date: Dec 2010
Publisher: [Honolulu] : [University of Hawaii at Manoa], [December 2010]
Abstract: This research focuses on the problem of job scheduling on homogeneous computational clusters. Clusters are widely used today for a variety of purposes, including high-performance scientific computing and Internet service hosting. While clusters may have impressive aggregate performance metrics, they are really only collections of fairly modest machines, which makes scheduling jobs for the best performance a non-trivial problem. Most clusters also need to be shared among users to amortize their start-up and maintenance costs, and ensuring that these users are treated fairly further adds to the difficulty. Existing approaches to scheduling attempt to address both of these issues, but have several limitations.
We propose a novel approach, called Dynamic Fractional Resource Scheduling (DFRS), to sharing homogeneous cluster computing platforms among competing jobs.
The key features of DFRS are that it leverages existing virtual machine technology in order to share resources more efficiently and it defines and optimizes a user-centric metric that captures notions of both performance and fairness. In this dissertation we explain the principles behind DFRS and its advantages over the current state of the art, develop a theoretical model of resource sharing, design heuristics to optimize the proposed metric within the given framework, implement and run simulations comparing DFRS to traditional approaches using popular and accepted performance metrics, and finally develop and test a prototype implementation based on existing technologies. Our results show that it is possible to develop heuristic algorithms that give results reasonably close to theoretical bounds for a variety of cases, that resource requirements are well within the capabilities of modern systems, and that for some scenarios DFRS can provide orders-ofmagnitude levels of improvement in performance over current approaches.
Description: Ph.D. University of Hawaii at Manoa 2010.
Includes bibliographical references.
Appears in Collections:Ph.D. - Computer Science

Items in ScholarSpace are protected by copyright, with all rights reserved, unless otherwise indicated.