Skip to Main content Skip to Navigation
Theses

Energy-aware management of scientific workflows in the Cloud : a Cloud provider-centric vision

Emile Cadorel 1, 2
Abstract : Scientific computer simulations are generally very complex and are characterized by many parallel processes. In order to highlight the parts that can be parallelized,and to enable efficient execution, many scientists have chosen to define their applications as workflows. A scientific workflow represents an application as a set of unitary processing tasks, linked by dependencies. Today, because of their low cost, elasticity, and on demand nature, cloud computing services are widely used for workflow execution. Users using this type of environment manage the execution of their workflow, as well as the necessary resources, using standard services such as IaaS (Infrastructure-as-a-Service). However,because cloud services are not specific to the nature of the application to be executed,the use of physical resources is not as optimized as it could be. In this thesis, we propose to move the management and execution of workflows to the cloud provider’s side in order to offer a new type of service dedicated to workflows. This new approach makes it possible to improve resource management and reduce energy consumption and thus the environmental impact of the infrastructure used.
Complete list of metadata

https://tel.archives-ouvertes.fr/tel-03248178
Contributor : Abes Star :  Contact
Submitted on : Thursday, June 3, 2021 - 2:22:08 PM
Last modification on : Saturday, June 5, 2021 - 3:17:06 AM

File

2020IMTA0195_Cadorel-Emile.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-03248178, version 1

Citation

Emile Cadorel. Energy-aware management of scientific workflows in the Cloud : a Cloud provider-centric vision. Distributed, Parallel, and Cluster Computing [cs.DC]. Ecole nationale supérieure Mines-Télécom Atlantique, 2020. English. ⟨NNT : 2020IMTA0195⟩. ⟨tel-03248178⟩

Share

Metrics

Record views

67

Files downloads

40