Performance implications of multi-tier application deployments on Infrastructure-as-a-Service clouds: Towards performance modeling
作者： W. Lloyd , S. Pallickara , O. David , J. Lyon , M. Arabi
DOI: 10.1016/J.FUTURE.2012.12.007
关键词:
摘要: Hosting a multi-tier application using an Infrastructure-as-a-Service (IaaS) cloud requires deploying components of the stack across virtual machines (VMs) to provide application's infrastructure while considering factors such as scalability, fault tolerance, performance and deployment costs (# VMs). This paper presents results from empirical study which investigates implications for resource requirements (CPU, disk network) resulting how applications are deployed IaaS clouds. We investigate of: (1) component placement VMs, (2) VM memory size, (3) hypervisor type (KVM vs. Xen), (4) physical hosts (provisioning variation). All possible configurations two variants tested. One variant was computationally bound by middleware, other geospatial queries. The best performing deployments required few 2 half number VM-level service isolation, demonstrating potential cost savings when can be consolidated. Resource utilization (CPU time, I/O, network I/O) varied with location, allocation, used (Xen or KVM) decisions impact resources. Isolating separate VMs produced overhead ~1%-2%. Provisioning variation up 3%. Relationships between were assessed multiple linear regression develop model predict performance. Our explained over 84% variance predicted mean absolute error only ~0.3 s CPU sector reads, writes serving most powerful predictors
colostate.edu 参考文章(29)
Malte Schwarzkopf, Steven Hand, Derek G. Murray, The seven deadly sins of cloud computing research usenix conference on hot topics in cloud ccomputing. pp. 1- 1 ,(2012)
Rini Mahajan, Dr. Dheerendra Singh, Cloud Computing Issues international conference on bioinformatics. ,vol. 4, pp. 626- 630 ,(2005) , 10.24297/IJCT.V4I2C2.4185
Matei Zaharia, Andy Konwinski, Anthony D. Joseph, Ion Stoica, Randy Katz, Improving MapReduce performance in heterogeneous environments operating systems design and implementation. pp. 29- 42 ,(2008) , 10.5555/1855741.1855744
Sriram Govindan, Jie Liu, Aman Kansal, Anand Sivasubramaniam, Cuanta: quantifying effects of shared on-chip resource interference for consolidated virtual machines symposium on cloud computing. pp. 22- ,(2011) , 10.1145/2038916.2038938
D. Armstrong, K. Djemame, Performance Issues in Clouds The Computer Journal. ,vol. 54, pp. 836- 849 ,(2011) , 10.1093/COMJNL/BXR011
Bikash Sharma, Victor Chudnovsky, Joseph L. Hellerstein, Rasekh Rifaat, Chita R. Das, Modeling and synthesizing task placement constraints in Google compute clusters symposium on cloud computing. pp. 3- ,(2011) , 10.1145/2038916.2038919
Thomas Voith, Karsten Oberle, Manuel Stein, Quality of service provisioning for distributed data center inter-connectivity enabled by network virtualization Future Generation Computer Systems. ,vol. 28, pp. 554- 562 ,(2012) , 10.1016/J.FUTURE.2011.03.011
George Kousiouris, Tommaso Cucinotta, Theodora Varvarigou, The effects of scheduling, workload type and consolidation scenarios on virtual machine performance and their prediction through optimized artificial neural networks Journal of Systems and Software. ,vol. 84, pp. 1270- 1291 ,(2011) , 10.1016/J.JSS.2011.04.013
Cheng-Zhong Xu, Jia Rao, Xiangping Bu, URL: A unified reinforcement learning approach for autonomic cloud management Journal of Parallel and Distributed Computing. ,vol. 72, pp. 95- 105 ,(2012) , 10.1016/J.JPDC.2011.10.003
M. Suhail Rehman, Majd F. Sakr, Initial Findings for Provisioning Variation in Cloud Computing ieee international conference on cloud computing technology and science. pp. 473- 479 ,(2010) , 10.1109/CLOUDCOM.2010.47