Run-Time QoS Optimisation

Run-Time Application QoS Optimisation in Private or Public Data Centre Clouds

The availability, load, and throughput of hardware resources (CPU, storage, and network) can vary in unpredictable ways, so ensuring that cloud-hosted applications achieve performance targets can be difficult. Studies have proved that performance uncertainty is the chief technical obstacle to successful adoption of cloud computing. The recent very high-profile crash of Amazon EC2 cloud, which took down the  applications of many SMEs, is a salient example of unpredictability in cloud environments. Some applications were down for hours, others for days. Theoretically, the elasticity provided by cloud computing can accommodate even unexpected changes in capacity, adding hardware resources when needed, and reducing them during periods of low demand, but the decisions to adjust capacity must be made frequently, automatically, and accurately to be cost effective.

The failure or congestion of network links are sometimes inevitable, given the scale, dynamics, and complexity in cloud computing, the crash or malfunction of a hardware resource, changes in workload patterns, or overloading of a hardware resource. Worse still, hardware resource status can be changed intentionally through malicious external interference. To deal with exceptions in cloud computing, several reactive techniques  have been proposed. These techniques rely on monitoring the state of hardware resources (which host instances of software resources) and triggering hard-coded, pre-configured corrective actions (e.g. allocating a new instance of software resource) when some event occurs (e.g. utilization of host CPU resource reaches a certain threshold). Reactive techniques are not sufficient to ensure guaranteed performance in the event of variation in state of cloud resources.

To develop techniques that can dynamically predict and capture the relationship between an application performance targets, current hardware resource allocation and changes in workload patterns, in order to adjust resource configuration remains an open research problem. Overall, the integration of theoretical workload prediction, resource performance models and optimization techniques to effect an end-to-end automated provisioning process over cloud resources is a hitherto neglected research area.

Further Reading

1. K. Alhamazani, R. Ranjan, P. Jayaraman, K. Mitra, F. Rabhi, D. Georgakopulos, and L. Wang,  “Cross-Layer Multi-Cloud Real-Time Application QoS Monitoring and Benchmarking As-a-Service Framework,” IEEE Transactions on Cloud Computing, IEEE Computer Society. (Accepted April 2015, in press)

2. K. Alhamazani, Rajiv Ranjan, F. Rabhi, L. Wang and K. Mitra, “Cloud Monitoring for Optimizing the QoS of Hosted Applications”, 4th IEEE International Conference on Cloud Computing Technology and Science, December 2012, 6 Pages, IEEE Computer Society.[ERA  Ranked]

3. R. N. Calheiros, R. Ranjan, and R. Buyya, “Virtual Machine Provisioning Based on Analytical Performance and QoS in Cloud Computing Environments“, 40th International Conference on Parallel Processing (ICPP 2011), Taipei, Taiwan, September 13-16, 2011, IEEE Computer Society. [ERA A Ranking]

4. R. Buyya, R. Ranjan, and R. N. Calheiros, “InterCloud: Utility-Oriented Federation of Cloud Computing Environments for Scaling of Application Services”. Proceedings of the 10th International Conference on Algorihtms and Architectures for Parallel Processing (ICA3PP 2010), Busan, Korea, March, 2010.[ERA B Ranking, Keynote Paper]

5. R. Ranjan and L. Zhao, “Peer-to-Peer Service Provisioning in Cloud Computing Environments”, Journal of Supercomputing, special issue of ICCSA2009 (Guest Editor: David Taniar), OCT 2011, 20 Pages, Springer Verlag, Invited Paper, In Press, doi: 10.1007/s11227-011-0710-5. [ERA B Ranking, ISI impact factor: 0.91]

6. R. Ranjan, R. Buyya, and M. Parashar, “Special Section on Autonomic Cloud Computing: Technologies, Services, and Applications”, In the Journal of Concurrency and Computation: Practice and Experience (CCPE), Volume 24, Issue 9, Pages 935–937, 25 June 2012, Wiley Press, doi: 10.1002/cpe.1865. [ERA A Ranked Journal, ISI impact factor: 0.84]  

7. M. Rahman, Md. Hassan, R. Ranjan, and R.Buyya, “Adaptive Workflow Scheduling in Dynamic Grid and Cloud Computing Environment”, Special Issue on Workflow Management in Service and Cloud Computing, In the Journal of Concurrency and Computation: Practice and Experience, Published online: 4 MAR 2013, DOI:  10.1002/cpe.3003, Wiley Press. [ERA A Ranked Journal, ISI impact factor: 0.84 ]      

8.  A. Guabtni, R. Ranjan, F. Rabhi, “A Workload-driven Approach to Database Query Processing in the Cloud”, In the Journal of Supercomputing, Volume 63, Issue 3, March 2013, Pages 722-736, Springer Netherlands, Press, doi: 10.1007/s11227-011-07. [ERA B Ranking, ISI impact factor: 0.91]

9. Zheng Li, Liam O’Brien, He Zhang, and Rajiv Ranjan, “Applying Design of Experiments (DOE) to Performance Evaluation of Commercial Cloud Services”, International Journal of Grid and High Performance Computing (IJGHPC), IGI Global Publisher, Accepted May 2013. [ACM Digital Library and DBLP Indexed]

10. C. Yang, J. Liu, R. Ranjan, W. Shih, and C. Lin, “On Construction of Heuristic QoS Bandwidth Management in Clouds”, In the Journal of Concurrency and Computation: Practice and Experience (CCPE), Wiley Press, Accepted: June  2013.  [ERA A Journal, ISI impact factor: 0.84]

11. R. Ranjan, R. Buyya, and S. Nepal, “Model-driven Provisioning of Application Services in Hybrid Computing Environments”, Future Generation Computer Systems Journal,  Volume 29, Issue 5, July 2013, Pages 1211–1215, Elsevier Press. [ERA A Journal, ISI impact factor 1.978]