Scheduling HPC Workflows for Responsiveness and Fairness with Networking Delays and Inaccurate Estimates of Execution Times
Andrew Burkimsher, Iain Bate and Leandro Soares Indrusiak
High-Performance Computing systems (HPCs) have grown in popularity in recent years, especially in the form of Grid and Cloud platforms. These platforms may be subject to periods of overload. In our previous research, we found that the Projected-SLR list scheduling policy provides responsiveness and a starvation-free scheduling guaran- tee in a realistic HPC scenario. This paper extends the previous work to consider networking delays in the platform model and inaccurate esti- mates of execution times in the application model. P-SLR is shown to be competitive with the best alternative scheduling policies in the presence of network costs (up to 400
BibTex Entry
@incollection{Burkimsher2013,
author = {Andrew Burkimsher and Iain Bate and Leandro Soares Indrusiak},
booktitle = {Proceedings of the 19th International Conference on Parallel Processing (Euro-Par 2013)},
doi = {10.1007/978-3-642-40047-6_15},
editor = {Felix Wolf and Bernd Mohr and Dieter Mey},
isbn = {978-3-642-40046-9},
link = {http://dx.doi.org/10.1007/978-3-642-40047-6_15},
pages = {126-137},
publisher = {Springer Berlin Heidelberg},
series = {Lecture Notes in Computer Science},
title = {Scheduling HPC Workflows for Responsiveness and Fairness with Networking Delays and Inaccurate Estimates of Execution Times},
volume = {8097},
year = {2013}
}