Performance Analysis for Large IaaS Clouds - Large Scale and Big Data: Processing and Management - page 576

In [6], Genaud et al. used simulations to evaluate scheduling strategies in cloud.
The authors primarily focused on minimizing the wait time of jobs and the monetary
cost of the rented resources. Our numeric-analytic models can complement such
studies. Goudarzi et al. [10] studied optimization problems for multitier cloud appli-
cations by taking into account requests with different CPU, memory, and network
resource requirements. Performance models developed in this paper can be extended
for such heterogeneous requests. Yigitbasi et al. [24] performed an experimental
analysis for the resource acquisition and release times in Amazon EC2. Our stochas-
tic models can complement these studies and analyze root-causes behind the major
results. Performance analysis of VM live migration and its impacts on SLAs was
studied by Voorsluys et al. [22]. Examples of other measurement based performance
evaluation of cloud services include [2,4,12,13,18].
18.6 CONCLUSIONS
With increasing popularity of IaaS cloud-based services, data-intensive applications
will find a natural home in such shared and virtualized environments. While a vari-
ety of cloud providers offer similar services, performance will be a key differentiator
among the delivered services. This chapter presents how model driven stochastic
performance analysis can scale for large sized clouds. The main idea is to develop
interacting submodels so that the overall model becomes scalable and tractable. Once
developed, such scalable models can be used for what-if analysis, bottleneck detec-
tion, and capacity planning. Tools and software packages can be developed based on
such models to assist a cloud administrator overseeing Big Data applications.
REFERENCES
1. Tutorial on the Sharpe interface, June 2013. http://sharpe.pratt.duke.edu/node/4.
2. A. Bhadani and S. Chaudhary. Performance evaluation of web servers using central load
balancing policy over virtual machines on cloud. In COMPUTE ' 10: Proceedings of the
Third Annual ACM Bangalore Conference , pp. 16:1-16:4, Bangalore, India, 2010.
3. R. Buyya, R. Ranjan, and R. N. Calheiros. Modeling and simulation of scalable cloud
computing environments and the cloudsim toolkit: Challenges and opportunities. In
IEEE International Conference on High Performance Computing and Simulation
(HPCS) , pp. 1-11, Leipzig, Germany, 2009.
4. J. Che, Q. He, K. Ye, and D. Huang. Performance Combinative Evaluation of Typical
Virtual Machine Monitors. In Second International Conference on High Performance
Computing and Applications (HPCA) , pp. 96-101, Shanghai, China, 2009.
5. G. Ciardo and K. S. Trivedi. A decomposition approach for stochastic reward net mod-
els. Elsevier Performance Evaluation , 18(1):37-59, 1993.
6. S. Genaud and J. Gossa. Cost-wait trade-offs in client-side resource provisioning with
elastic clouds. In IEEE International Conference on Cloud Computing (CLOUD) ,
pp. 1-8, Washington, DC, July 2011.
7. R. Ghosh. Scalable stochastic models for cloud services. PhD Thesis, Duke University ,
2012.
8. R. Ghosh, F. Longo, V. K. Naik, and K. S. Trivedi. Modeling and performance analy-
sis of large scale IaaS clouds. Elsevier Future Generation Computer Systems , 2012.
Available online: http://dx.doi.org/10.1016/j.future.2012.06.005.