DSpace About DSpace Software
 

Makerere University Research Repository >
Faculty of Computing and Information Technology >
Research Articles (CIT) >

Please use this identifier to cite or link to this item: http://hdl.handle.net/123456789/493

Title: The greedy multi-cluster scheduler: performance bounds and parametric sensitivity
Authors: Ngubiri, John
van Vliet, Mario
Keywords: Greedy
Scheduler
Co-allocation
Multi-cluster
Issue Date: 2007
Publisher: IJCIR
Citation: International Journal of Computing and ICT Research
Abstract: Most schedulers in parallel job scheduling do not put (job)schedulability into consideration when prioritizing jobs. Performance evaluation is mostly done using average values of the measurement metric. Using the average metric value may conceal relative job starvation details hence giving a shallower understanding of scheduler performance. We propose a greedy multi-cluster scheduler that uses the (estimate of) job schedulability and the time a job has spent in the queue to compute its priority.We compare the performance of our scheduler with that of Fit Processor First Served (FPFS) scheduler. We also study the sensitivity of its performance to parameter changes. We observe that (i) within some parameter ranges, our scheduler outperforms FPFS; (ii) for big jobs, our scheduler outperforms FPFS; for small jobs, FPFS outperforms our scheduler and (iii) our scheduler is fairer than FPFS.
Description: The paper proposes and evaluates a new co-allocation algorithm
URI: http://hdl.handle.net/123456789/493
ISSN: 1996-1065
Appears in Collections:Research Articles (CIT)

Files in This Item:

File Description SizeFormat
ngubiri_vanVliet_ijcir.PDF1207KbAdobe PDFView/Open

All items in DSpace are protected by copyright, with all rights reserved.

 

Valid XHTML 1.0! DSpace Software Copyright © 2002-2005 MIT and Hewlett-Packard - Feedback