Budget-aware scheduling algorithms for scientific workflows on IaaS cloud platforms

Yves Caniou 1, 2 Eddy Caron 1, 2 Aurélie Kong Win Chang 1, 3, 2 Yves Robert 1, 3
2 AVALON - Algorithms and Software Architectures for Distributed and HPC Platforms
Inria Grenoble - Rhône-Alpes, LIP - Laboratoire de l'Informatique du Parallélisme
3 ROMA - Optimisation des ressources : modèles, algorithmes et ordonnancement
Inria Grenoble - Rhône-Alpes, LIP - Laboratoire de l'Informatique du Parallélisme
Abstract : This report introduces several budget-aware algorithms to deploy scientific workflows on IaaS cloud platforms, where users can request Virtual Machines (VMs) of different types, each with specific cost and speed parameters. We use a realistic application/platform model with stochastic task weights, and VMs communicating through a datacenter. We extend two well-known algorithms, HEFT and \MinMin, and make scheduling decisions based upon machine availability and available budget. During the mapping process, the budget-aware algorithms make conservative assumptions to avoid exceeding the initial budget; we further improve our results with refined versions that aim at re-scheduling some tasks onto faster VMs, thereby spending any budget fraction leftover by the first allocation. These refined variants are much more time-consuming than the former algorithms, so there is a trade-off to find in terms of scalability. We report an extensive set of simulations with workflows from the Pegasus benchmark suite. Budget-aware algorithms generally succeed in achieving efficient makespans while enforcing the given budget, and despite the uncertainty in task weights.
Document type :
Reports
Liste complète des métadonnées

https://hal.inria.fr/hal-01574491
Contributor : Equipe Roma <>
Submitted on : Monday, August 14, 2017 - 5:31:08 PM
Last modification on : Friday, April 20, 2018 - 5:01:52 PM

File

RR-9088.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01574491, version 1

Citation

Yves Caniou, Eddy Caron, Aurélie Kong Win Chang, Yves Robert. Budget-aware scheduling algorithms for scientific workflows on IaaS cloud platforms. [Research Report] RR-9088, INRIA. 2017, pp.27. ⟨hal-01574491⟩

Share

Metrics

Record views

192

Files downloads

201