The automatic allocation of enterprise workload to resources can be enhanced by being able to make what-if response time predictions whilst different allocations are being considered. We experimentally investigate an historical and a layered queuing performance model and show how they can provide a good level of support for a dynamic-urgent cloud environment. Using this we define, implement and experimentally investigate the effectiveness of a prediction-based cloud workload and resource management algorithm. Based on these experimental analyses we: (i) comparatively evaluate the layered queuing and historical techniques; (ii) evaluate the effectiveness of the management algorithm in different operating scenarios; and (iii) provide guidance on using prediction-based workload and resource management.
|Number of pages||17|
|Journal||Simulation Modelling Practice and Theory|
|Publication status||Published - Jun 2011|
Bibliographical noteFunding Information:
This work was sponsored in part by the JISC TeciRes project, the Technology Strategy Board FireGrid Project, The University of Edinburgh School of Informatics, the EPSRC DOPCHE Project (GR/S03058/01). We gratefully acknowledge the assistance during current and earlier versions of this work of: M. Atkinson, G. Beckett, J. Koetsier, S. Koo, T. Liu, I. Mitrani, A. van Moorsel, G. Nudd, S. Potter, G. Pringle, J. Slegers, C. Smith, D. Spooner, N. Thomas, J. Torero, S. Welch, W. J. Xue, S. J. Pennycook and all FireGrid team members not mentioned here.
- HYDRA historical model
- Layered queuing
- Performance modelling
ASJC Scopus subject areas
- Modelling and Simulation
- Hardware and Architecture