A Heuristic Scheduling and Resource Management System for Solving Bioinformatical Problems via High Performance Computing on Heterogeneous Multi-platform Hardware
To process the data available in Bioinformatics, High Performance Computing is required. To efficiently calculate the necessary data, the computational tasks need to be scheduled and maintained. We propose a method of predicting runtimes in a heterogeneous high performance computing environment as well as scheduling methods for the execution of hgih performance tasks. The heuristic method used is the feedforward artificial neural network, which utilizes a collected history of real life data to predict and schedule upcoming jobs.
KeywordsPrediction Error High Performance Computing Schedule System Heuristic Schedule Node Hide Layer
Unable to display preview. Download preview PDF.
- 1.Condor Team: Condor Version 7.4.2 Manual. University of Wisconsin-Madison (May 2010)Google Scholar
- 2.SLURM Team: SLURM: A Highly Scalable Resource Manager. SLURM Website (June 2010) (last visit July 15, 2010)Google Scholar
- 3.Templeton, D.: A Beginner’s Guide to Sun Grid Engine 6.2. Whitepaper, Sun Microsystems (July 2009)Google Scholar