2016 24th Euromicro International Conference on Parallel, Distributed, and Network-Based Processing (PDP) (2016)
Heraklion, Crete, Greece
Feb. 17, 2016 to Feb. 19, 2016
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/PDP.2016.39
Day after day, cloud technologies are more and more adopted by very diverse types of stakeholders, and this success creates a side-effect problem: the energy spent by this kind of infrastructures is growing bigger every day. With the objective of reducing energy consumption when programming applications for cloud infrastructures, we have implemented energy-aware mechanisms in the COMPSs Programming Model, inside the context of the ASCETiC Project. In this paper, we demonstrate that application-level scheduling can have a big impact on the energy consumed by an application when executed in a heterogeneous cloud. We have implemented an energy-aware scheduling mechanism in COMPSs, together with a versioning technique, and we have run experiments with a use case coming from the real estate sector that proves our hypotheses.
Energy consumption, Runtime, Cloud computing, Programming profession, Processor scheduling, Green products
F. Lordan, J. Ejarque, R. Sirvent and R. M. Badia, "Energy-Aware Programming Model for Distributed Infrastructures," 2016 24th Euromicro International Conference on Parallel, Distributed, and Network-Based Processing (PDP), Heraklion, Crete, Greece, 2016, pp. 413-417.