ABSTRACT
Infrastructure-as-a-system compute clouds such as Amazon's EC2 allow users to pay a flat hourly rate to run their virtual machine (VM) on a server providing some combination of CPU access, storage, and network. But not all VM instances are created equal: distinct underlying hardware differences, contention, and other phenomena can result in vastly differing performance across supposedly equivalent instances. The result is striking variability in the resources received for the same price.
We initiate the study of customer-controlled placement gaming: strategies by which customers exploit performance heterogeneity to lower their costs. We start with a measurement study of Amazon EC2. It confirms the (oft-reported) performance differences between supposedly identical instances, and leads us to identify fruitful targets for placement gaming, such as CPU, network, and storage performance. We then explore simple heterogeneity-aware placement strategies that seek out better-performing instances. Our strategies require no assistance from the cloud provider and are therefore immediately deployable. We develop a formal model for placement strategies and evaluate potential strategies via simulation. Finally, we verify the efficacy of our strategies by implementing them on EC2; our experiments show performance improvements of 5% for a real-world CPU-bound job and 34% for a bandwidth-intensive job.
- Amazon Ltd. Amazon elastic compute cloud (EC2). http://aws.amazon.com/ec2/.Google Scholar
- Amazon Web Services. Amazon EC2 instance types. http://aws.amazon.com/ec2/instance-types/.Google Scholar
- Amazon Web Services. Amazon Elastic Block Store (EBS). http://aws.amazon.com/ebs/.Google Scholar
- Apache Software Foundation. ab - apache http server benchmarking tool. http://httpd.apache.org/docs/2.0/programs/ab.html.Google Scholar
- Apache Software Foundation. Apache HTTP server project. http://httpd.apache.org/.Google Scholar
- Manjari Asawa and Demosthenis Teneketzis. Multi-armed bandits with switching penalties. IEEE Transactions on Automatic Control, 41(3), March 1996.Google ScholarCross Ref
- Sean Kenneth Barker and Prashant Shenoy. Empirical evaluation of latency-sensitive application performance in the cloud. In MMsys, pages 35--46, 2010. Google ScholarDigital Library
- Russker Coker. Bonnie++ benchmark version 1.03e. http://www.coker.com.au/bonnie++/, 2008.Google Scholar
- Jiang Dejun, Guillaume Pierre, and Chi-Hung Chi. Ec2 performance analysis for resource provisioning of service-oriented applications. In Proceedings of the 2009 international conference on Service-oriented computing, pages 197--207, 2009. Google ScholarDigital Library
- Jiang Dejun, Guillaume Pierre, and Chi-Hung Chi. Resource provisioning of web applications in heterogeneous clouds. In Proceedings of the 2nd USENIX conference on Web application development, pages 5--5, 2011. Google ScholarDigital Library
- Diwaker Gupta, Ludmila Cherkasova, Rob Gardner, and Amin Vahdat. Enforcing performance isolation across virtual machines in xen. In Middleware, 2006. Google ScholarDigital Library
- J. L. Henning. Spec cpu2006 benchmark descriptions. In SIGARCH Computer Architecture News, 2006. Google ScholarDigital Library
- Alexandru Iosup, Nezih Yigitbasi, and Dick H. J. Epema. On the performance variability of production cloud services. In 11th IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID), pages 104--113, 2011. Google ScholarDigital Library
- Aamer Jaleel, Hashem H. Najaf-abadi, Samantika Subramaniam, Simon C. Steely, and Joel Emer. Cruise: cache replacement and utility-aware scheduling. In ASPLOS, 2012. Google ScholarDigital Library
- Vivek Kale, Jayanta Mukherjee, and Indranil Gupta. Hadoopjitter: The ghost in the machine and how to tame it. http://hdl.handle.net/2142/17084, 2010.Google Scholar
- Yaakoub El Khamra, Hyunjoo Kim, Shantenu Jha, and Manish Parashar. Exploring the performance fluctuations of hpc workloads on clouds. In CloudCom, pages 383--387, 2010. Google ScholarDigital Library
- Dan Klein, Joseph Smarr, Huy Nguyen, and Christopher D. Manning. Named entity recognition with character-level models. In Proceedings the Seventh Conference on Natural Language Learning, pages 180--183, 2003. Google ScholarDigital Library
- A. Li, X. Yang, S. Kandula, and M. Zhang. Cloudcmp: Comparing public cloud providers. In IMC, 2010. Google ScholarDigital Library
- Seung-Hwan Lim, Jae-Seok Huh, Youngjae Kim, and Chita R. Das. Migration, assignment, and scheduling of jobs in virtualized environment. In HotCloud, 2010. Google ScholarDigital Library
- Dave Mangot. EC2 variability: The numbers revealed. http://tech.mangot.com/roller/dave/entry/ec2_variability_the_numbers_re%vealed, May 2009.Google Scholar
- Microsoft Corp. Windows azure: Pricing details. http://www.windowsazure.com/en-us/pricing/details/.Google Scholar
- Zhonghong Ou, Hao Zhuang, Jukka K. Nurminen, Antti Ylä-Jääski, and Pan Hui. Exploiting hardware heterogeneity within the same instance type of amazon EC2. In 4th USENIX Workshop on Hot Topics in Cloud Computing (HotCloud), 2012. Google ScholarDigital Library
- R. S. Sutton and A. G. Barto. Reinforcement Learning: An Introduction. MIT Press, 1998. Google ScholarDigital Library
- Rackspace Inc. How we price cloud servers. http://www.rackspace.com/cloud/cloud_hosting_products/servers/pricing/.Google Scholar
- M. S. Rehman and M. F. Sakr. Initial findings for provisioning variation in cloud computing. In IEEE Second International Conference on Cloud Computing Technology and Science (CloudCom), pages 473--479, 2010. Google ScholarDigital Library
- T. Ristenpart, E. Tromer, H. Shacham, and S. Savage. Hey, you, get off my cloud: exploring information leakage in third party compute clouds. In CCS, 2009. Google ScholarDigital Library
- Juan Carlos Saez, Manuel Prieto, Alexandra Fedorova, and Sergey Blagodurov. A comprehensive scheduler for asymmetric multicore systems. In Proc. EuroSys, 2010. Google ScholarDigital Library
- Jorg Schad, Jens Dittrich, and Jorge-Arnulfo Quiané-Ruiz. Runtime measurements in the cloud: Observing, analyzing, and reducing variance. In VLDB, September 2010. Google ScholarDigital Library
- Malte Schwarzkopf, Derek G. Murray, and Steven Hand. The seven deadly sins of cloud computing research. In HotCloud, June 2012. Google ScholarDigital Library
- Burr Settles. Biomedical named entity recognition using conditional random fields and rich feature sets. In NLPBA, 2004. Google ScholarDigital Library
- Alan Shieh, Srikanth Kandula, Albert Greenberg, and Changhoon Kim. Seawall: Performance isolation for cloud datacenter networks. In HotCloud, 2010. Google ScholarDigital Library
- The Stanford Natural Language Processing Group. Stanford named entity recognizer (NER) version 1.2.4. http://nlp.stanford.edu/software/CRF-NER.shtml, 2012.Google Scholar
- A. Tirumala, F. Qin, J Dugan, J. Ferguson, and K. Gibbs. Iperf: The TCP/UDP bandwidth measurement tool, version 2.0.5. http://sourceforge.net/projects/iperf/, 2010.Google Scholar
- Guohui Wang and T. S. Eugene Ng. The impact of virtualization on network performance of amazon EC2 data center. In IEEE INFOCOM, 2010. Google ScholarDigital Library
- P. Whittle. Sequential scheduling and the multi-armed bandit, chapter 14. In Optimization over Time--Dynamic Programming and Stechastic Control. vol. 1, pages 210--219. John Wiley and Sons Ltd., 1982.Google Scholar
- Wikipedia. Markov decision process. http://en.wikipedia.org/wiki/Markov_decision_process.Google Scholar
- J. Xie, S. Yin, X.-J. Ruan, Z.-Y. Ding, Y. Tian, J. Majors, and X. Qin. Improving MapReduce performance via data placement in heterogeneous Hadoop Clusters. In Proc. 19th Int'l Heterogeneity in Computing Workshop, April 2010.Google Scholar
- Jun-Ming Xu, Kwang-Sung Jun, Xiaojin Zhu, and Amy Bellmore. Learning from bullying traces in social media. In NAACL HLT, 2012. Google ScholarDigital Library
- Sungkap Yeo and Hsien-Hsin S. Lee. Using mathematical modeling in provisioning a heterogeneous cloud computing environment. IEEE Computer, 44(8), August 2011. Google ScholarDigital Library
- Matei Zaharia, Andy Konwinski, Anthony D. Joseph, Randy Katz, and Ion Stoica. Improving mapreduce performance in heterogeneous environments. In OSDI, pages 29--42, 2008. Google ScholarDigital Library
- Ce Zhang, Feng Niu, Christopher Ré, and Jude Shavlik. Big data versus the crowd: Looking for relationships in all the right places. In ACL, 2012. Google ScholarDigital Library
Index Terms
More for your money: exploiting performance heterogeneity in public clouds
Recommendations
Automatic Consolidation of Virtual Machines in On-Premises Cloud Platforms
CCGrid '17: Proceedings of the 17th IEEE/ACM International Symposium on Cluster, Cloud and Grid ComputingAfter a sequence of creation and destruction of virtual machines (VMs) in an on-premises Cloud computing platform, the scheduling decisions to host the VMs are far from being optimal and the fragmentation of the physical resources may impede the ...
A critical survey of live virtual machine migration techniques
Virtualization techniques effectively handle the growing demand for computing, storage, and communication resources in large-scale Cloud Data Centers (CDC). It helps to achieve different resource management objectives like load balancing, online system ...
Efficient Migration of Virtual Machines between Public and Private Cloud
CLOUDCOM '11: Proceedings of the 2011 IEEE Third International Conference on Cloud Computing Technology and ScienceCloud computing service providers offer cost-effective means to burst computational needs and utilise live migration of virtual machines (VMs) for effective and efficient work-load movements with short service downtimes. However, there is a lack of ...
Comments