loading...
Fair Load-Balancing on Parallel Systems for QoS
Xi'an, China September 10-September 14
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/ICPP.2007.412007 International Conference on Para ...
 This Article 
 
PDF
HTML
 
 Share 
   
 Bibliographic References 
   
 Add to: 
 
Digg
Furl
Spurl
Blink
Simpy
Google
Del.icio.us
Y!MyWeb
 
 Search 
   
L.F. Orleans, University of Coimbra, Portugal
P.N. Furtado, University of Coimbra, Portugal
Many of the load-balancing algorithms used in parallel systems do not have a concern about response times: tasks (or requests) are simply dispatched to a server, which provides no guarantees about their execution times. When there is a maximum acceptable response time (i.e. deadline) for tasks to be executed, the consequences caused by the adoption of traditional algorithms for load-balancing can be catastrophic: when the system is under heavy loads, a huge amount of tasks miss their deadlines, even the faster ones. Also, the number of longer tasks that ends is very small -- near to zero in all cases. In this paper we discuss why the traditional algorithms fail to provide the intended QoS capacity. Then, we present a new algorithm, "On-demand Restriction for Big Tasks (ORBITA)", which is proved, by simulation, to be a fair alternative for stressed systems since tasks of all durations have a chance to complete their execution before their deadlines are reached.
Citation:
L.F. Orleans, P.N. Furtado, "Fair Load-Balancing on Parallel Systems for QoS," icpp, pp.22, 2007 International Conference on Parallel Processing (ICPP 2007), 2007
Usage of this product signifies your acceptance of the Terms of Use.