Autonomous Resource-Aware Scheduling of Large-Scale Media Workflows

Abstract : The media processing and distribution industry generally requires considerable resources to be able to execute the various tasks and workflows that constitute their business processes. The latter processes are often tied to critical constraints such as strict deadlines. A key issue herein is how to efficiently use the available computational, storage and network resources to be able to cope with the high work load. Optimizing resource usage is not only vital to scalability, but also to the level of QoS (e.g. responsiveness or prioritization) that can be provided. We designed an autonomous platform for scheduling and workflow-to-resource assignment, taking into account the different requirements and constraints. This paper presents the workflow scheduling algorithms, which consider the state and characteristics of the resources (computational, network and storage). The performance of these algorithms is presented in detail in the context of a European media processing and distribution use-case.
Document type :
Conference papers
Complete list of metadatas

Cited literature [12 references]  Display  Hide  Download

https://hal.inria.fr/hal-01056624
Contributor : Hal Ifip <>
Submitted on : Wednesday, August 20, 2014 - 12:27:01 PM
Last modification on : Tuesday, October 17, 2017 - 11:40:31 AM
Long-term archiving on : Thursday, November 27, 2014 - 11:35:23 AM

File

61550051.pdf
Files produced by the author(s)

Licence


Distributed under a Creative Commons Attribution 4.0 International License

Identifiers

Citation

Stein Desmet, Bruno Volckaert, Filip Turck. Autonomous Resource-Aware Scheduling of Large-Scale Media Workflows. 4th International Conference on Autonomous Infrastructure, Management and Security (AIMS), Jun 2010, Zurich, Switzerland. pp.50-64, ⟨10.1007/978-3-642-13986-4_6⟩. ⟨hal-01056624⟩

Share

Metrics

Record views

84

Files downloads

116