Design and Evaluation of a Portfolio Scheduler for Business-Critical Workloads Hosted in Cloud Datacenters

Master thesis (2015)

Authors

Contributors

A. Iosup (mentor)
Programme
Computer Science (EEMCS) (TU Delft)
Copyright: © 2015 Van Beek, V.S.
More Info
expand_more

Published Date

27-01-2015

Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Programme

Computer Science

Abstract

Virtualized multi-cluster, multi-datacenter datacenters are central to the digital economy, but require new techniques to address increasing scales, changing architectures, and dynamic workloads. Especially the introduction of new Business-Critical workloads that we characterize, create new resource management challenges. We propose a resource management and scheduling architecture that uses resources efficiently and reduces the risk of low performance. For this we create and evaluate a portfolio scheduler that is adapted to support multi-cluster, multi-datacenter architectures.