State-Space Feedback Control for Elastic Distributed Storage in a Cloud Environment
2012 (English)In: ICAS 2012: The Eighth International Conference on Autonomic and Autonomous Systems, St. Maarten, Netherlands Antilles, 2012, 589-596 p.Conference paper (Refereed)
Elasticity in Cloud computing is an ability of asystem to scale up and down (request and release resources) in response to changes in its environment and workload. Elasticity can be achieved manually or automatically. Efforts arebeing made to automate elasticity in order to improve system performance under dynamic workloads. In this paper, we reportour experience in designing an elasticity controller for a key-value storage service deployed in a Cloud environment. To design our controller, we have adopted a control theoretic approach. Automation of elasticity is achieved by providing a feedback controller that automatically increases and decreases the number of nodes in order to meet service level objectives under high load and to reduce costs under low load. Every step in the building of a controller for elastic storage, includingsystem identification and controller design, is discussed. We have evaluated our approach by using simulation. We have developed a simulation framework EStoreSim in order to simulate anelastic key-value store in a Cloud environment and be able to experiment with different controllers. We have examined the implemented controller against specific service level objectives and evaluated the controller behavior in different scenarios. Our simulation experiments have shown the feasibility of our approach to automate elasticity of storage services using state-space feedback control.
Place, publisher, year, edition, pages
St. Maarten, Netherlands Antilles, 2012. 589-596 p.
elasticity, key-value store, Cloud, state-space feedback control
IdentifiersURN: urn:nbn:se:kth:diva-101375ISBN: 978-1-61208-187-8OAI: oai:DiVA.org:kth-101375DiVA: diva2:547225
The 8th International Conference on Autonomic and Autonomous Systems (ICAS 2012)
FunderICT - The Next Generation
QC 201512162012-08-272012-08-272015-12-16Bibliographically approved