Towards Scheduling Evolving Applications
Most high-performance computing resource managers only allow applications to request a static allocation of resources. However, evolving applications have resource requirements which change (evolve) during their execution. Currently, such applications are forced to make an allocation based on their peak resource requirements, which leads to an inefficient resource usage. This paper studies whether it makes sense for resource managers to support evolving applications. It focuses on scheduling fully-predictably evolving applications on homogeneous resources, for which it proposes several algorithms and evaluates them based on simulations. Results show that resource usage and application response time can be significantly improved with short scheduling times.
Unable to display preview. Download preview PDF.
- 3.Mu’alem, A.W., Feitelson, D.G.: Utilization, predictability, workloads, and user runtime estimates in scheduling the IBM SP2 with backfilling. TPDS 12(6) (2001)Google Scholar
- 4.Plewa, T., Linde, T., Weirs, V.G. (eds.): Adaptive Mesh Refinement – Theory and Applications. Springer (2003)Google Scholar
- 6.Ribes, A., Caremoli, C.: Salome platform component model for numerical simulation. COMPSAC 2, 553–564 (2007)Google Scholar
- 7.Hungershofer, J.: On the combined scheduling of malleable and rigid jobs. In: SBAC-PAD (2004)Google Scholar
- 8.Buisson, J., Sonmez, O., Mohamed, H., et al.: Scheduling malleable applications in multicluster systems. Technical Report TR-0092, CoreGRID (2007)Google Scholar
- 9.El Maghraoui, K., Desell, T.J., Szymanski, B.K., Varela, C.A.: Dynamic malleability in iterative MPI applications. In: CCGRID (2007)Google Scholar
- 10.Cera, M.C., Georgiou, Y., Richard, O., Maillard, N., Navaux, P.O.A.: Supporting MPI malleable applications upon the OAR resource manager. In: COLIBRI (2009) Google Scholar
- 11.Buisson, J., Sonmez, O., Mohamed, H., et al.: Scheduling malleable applications in multicluster systems. Technical Report TR-0092, CoreGRID (2007)Google Scholar
- 12.Adaptive Computing Enterprises, Inc.: Moab workload manager administrator guide, version 6.0.2, http://www.adaptivecomputing.com/resources/docs/mwm
- 13.Cycles, C.: Lessons learned building a 4096-core cloud HPC supercomputer, http://blog.cyclecomputing.com/2011/03/cyclecloud-4096-core-cluster.html