Abstract
Cloud computing promises flexibility and high performance for users and cost efficiency for operators. To achieve this, cloud providers offer instances of different sizes, both as long-term reservations and short-term, on-demand allocations. Unfortunately, determining the best provisioning strategy is a complex, multi-dimensional problem that depends on the load fluctuation and duration of incoming jobs, and the performance unpredictability and cost of resources. We first compare the two main provisioning strategies (reserved and on-demand resources) on Google Compute Engine (GCE) using three representative workload scenarios with batch and latency-critical applications. We show that either approach is suboptimal for performance or cost. We then present HCloud, a hybrid provisioning system that uses both reserved and on-demand resources. HCloud determines which jobs should be mapped to reserved versus on-demand resources based on overall load, and resource unpredictability. It also determines the optimal instance size an application needs to satisfy its Quality of Service (QoS) constraints. We demonstrate that hybrid configurations improve performance by 2.1x compared to fully on-demand provisioning, and reduce cost by 46% compared to fully reserved systems. We also show that hybrid strategies are robust to variation in system and job parameters, such as cost and system load.
- Amazon ec2. http://aws.amazon.com/ec2/.Google Scholar
- K. Annapureddy. Security challenges in hybrid cloud infrastructures. In Aalto University, T-110.5290 Seminar on Network Security. 2010.Google Scholar
- Autoscale. https://cwiki.apache.org/cloudstack/autoscaling.html.Google Scholar
- Aws autoscaling. http://aws.amazon.com/autoscaling/.Google Scholar
- G. Banga, P. Druschel, and J. Mogul. Resource containers: a new facility for resource management in server systems. In Proceeedings of OSDI. New Orleans, 1999.Google Scholar
- S. K. Barker and P. Shenoy. Empirical evaluation of latency-sensitive application performance in the cloud. In Proceedings of the First Annual ACM SIGMM Conference on Multimedia Systems (MMsys). Scottsdale, AR, 2010.Google Scholar
Digital Library
- L. Barroso. Warehouse-scale computing: Entering the teenage decade. ISCA Keynote, SJ, June 2011.Google Scholar
- L. Barroso and U. Hoelzle. The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines. MC Publishers, 2009.Google Scholar
Digital Library
- O. A. Ben-Yehuda, M. Ben-Yehuda, A. Schuster, and D. Tsafrir. Deconstructing amazon ec2 spot instance pricing. In ACM TEAC, 1(3), September 2013.Google Scholar
- G. Breiter and V. Naik. A framework for controlling and managing hybrid cloud service integration. In Proceedings of the 2013 IEEE International Conference on Cloud Engineering (IC2E). Redwood City, CA, 2013.Google Scholar
Digital Library
- M. Carvalho, W. Cirne, F. Brasileiro, and J. Wilkes. Long-term slos for reclaimed cloud computing resources. In Proceedings of ACM Symposium on Cloud Computing (SOCC). Seattle, WA, 2014.Google Scholar
Digital Library
- J. Chase, D. Anderson, P. Thakar, A. Vahdat, and R. Doyle. Managing energy and server resources in hosting centers. In Proceedings of the Symposium on Operating Systems Principles (SOSP). Banff, CA, 2001.Google Scholar
Digital Library
- Linux containers. http://lxc.sourceforge.net/.Google Scholar
- J. Dean and L. A. Barroso. The tail at scale. In Communications of the ACM (CACM), Vol. 56 No. 2, Pages 74--80.Google Scholar
- E. Deelman, G. Singh, M. Livny, B. Berriman, and J. Good. The cost of doing science on the cloud: The montage example. In Proceedings of Supercomputing (SC). Austin, TX, 2008.Google Scholar
Cross Ref
- C. Delimitrou, N. Bambos, and C. Kozyrakis. QoS-Aware Admission Control in Heterogeneous Datacenters. In Proceedings of the International Conference of Autonomic Computing (ICAC). San Jose, CA, USA, 2013.Google Scholar
- C. Delimitrou and C. Kozyrakis. iBench: Quantifying Interference for Datacenter Workloads. In Proceedings of the IEEE International Symposium on Workload Characterization (IISWC). Portland, OR, September 2013.Google Scholar
- C. Delimitrou and C. Kozyrakis. Paragon: QoS-Aware Scheduling for Heterogeneous Datacenters. In Proceedings of the Eighteenth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS). Houston, TX, 2013.Google Scholar
Digital Library
- C. Delimitrou and C. Kozyrakis. QoS-Aware Scheduling in Heterogeneous Datacenters with Paragon. In ACM Transactions on Computer Systems (TOCS), Vol. 31 Issue 4. December 2013.Google Scholar
Digital Library
- C. Delimitrou and C. Kozyrakis. Quality-of-Service-Aware Scheduling in Heterogeneous Datacenters with Paragon. In IEEE Micro Special Issue on Top Picks from the Computer Architecture Conferences. May/June 2014.Google Scholar
- C. Delimitrou and C. Kozyrakis. Quasar: Resource-Efficient and QoS-Aware Cluster Management. In Proceedings of the Nineteenth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS). Salt Lake City, UT, USA, 2014.Google Scholar
Digital Library
- C. Delimitrou, D. Sanchez, and C. Kozyrakis. Tarcil: Reconciling Scheduling Speed and Quality in Large Shared Clusters. In Proceedings of the Sixth ACM Symposium on Cloud Computing (SOCC). Kohala Coast, HI, 2015.Google Scholar
Digital Library
- B. Farley, A. Juels, V. Varadarajan, T. Ristenpart, K. D. Bowers, and M. M. Swift. More for your money: Exploiting performance heterogeneity in public clouds. In Proceedings of the ACM Symposium on Cloud Computing (SOCC). San Jose, CA, 2012.Google Scholar
Digital Library
- Google compute engine. https://developers.google.com/compute/.Google Scholar
- D. Gmach, J. Rolia, L. Cherkasova, and A. Kemper. Workload analysis and demand prediction of enterprise data center applications. In Proceedings of the IEEE International Symposium on Workload Characterization (IISWC). Boston, MA, 2007.Google Scholar
Digital Library
- Z. Gong, X. Gu, and J. Wilkes. Press: Predictive elastic resource scaling for cloud systems. In Proceedings of the International Conference on Network and Service Management (CNSM). Niagara Falls, ON, 2010.Google Scholar
- G. R. Grimmett and D. R. Stirzaker. Probability and random processes. 2nd Edition. Clarendon Press, Oxford, 1992.Google Scholar
- M. Guevara, B. Lubin, and B. Lee. Navigating heterogeneous processors with market mechanisms. In Proceedings of the IEEE Symposium on High Performance Computer Architecture (HPCA). Shenzhen, China, 2013.Google Scholar
Digital Library
- J. Hamilton. Cost of power in large-scale data centers. http://perspectives.mvdirona.com.Google Scholar
- B. Hindman, A. Konwinski, M. Zaharia, A. Ghodsi, A. Joseph, R. Katz, S. Shenker, and I. Stoica. Mesos: A platform for fine-grained resource sharing in the data center. In Proceedings of the USENIX Sumposium on Networked Systems Design and Implementation (NSDI). Boston, MA, 2011.Google Scholar
- M. Hoseinyfarahabady, H. Samani, L. Leslie, Y. C. Lee, and A. Zomaya. Handling uncertainty: Pareto-efficient bot scheduling on hybrid clouds. In Proceedings of the International Conference for Parallel Processing (ICPP). Lyon, France, 2013.Google Scholar
Digital Library
- A. Iosup, N. Yigitbasi, and D. Epema. On the performance variability of production cloud services. In Proceedings of IEEE/ACM International Symposium on Cluster, Cloud and Grid Computing (CCGRID). Newport Beach, CA, 2011.Google Scholar
Digital Library
- V. Khadilkar, K. Y. Oktay, B. Hore, M. Kantarcioglu, S. Mehrotra, and B. Thuraisingham. Risk-aware data processing in hybrid clouds. TR-UTDCS-31--11, 2011.Google Scholar
- Y. E. Khamra, H. Kim, S. Jha, and M. Parashar. Exploring the performance fluctuations of hpc workloads on clouds. In Proceedings of IEEE International Conference on Cloud Computing Technology and Science (CloudCom). Indianapolis, IN, 2010.Google Scholar
- L. Kleinrock. Queueing systems volume 1: Theory. pp. 101--103, 404.Google Scholar
- E. Lau, J. E. Miller, I. Choi, D. Yeung, S. Amarasinghe, and A. Agarwal. Multicore performance optimization using partner cores. In Proceedings of the USENIX Workshop on Hot Topics in Parallelism (HotPar). Berkeley, CA, 2011.Google Scholar
- J. Laudon. Performance/watt: the new server focus. In ACM SIGARCH Computer Architecture News: dasCMP. Vol. 33 Issue 4, p. 5--13, November 2005.Google Scholar
Digital Library
- A. Li, X. Yang, S. Kandula, and M. Zhang. Cloudcmp: Comparing public cloud providers. In Proceedings of Internet Measurement Conference (IMC). Melbourne, Australia, 2010.Google Scholar
Digital Library
- Host server cpu utilization in amazon ec2 cloud. http://goo.gl/nCfYFX.Google Scholar
- J. Lorch and A. Smith. Reducing processor power consumption by improving processor time management in a single-user operating system. In Proceedings of Annual International Conference on Mobile Computing and Networking (Mobicom). New York, NY, 1996.Google Scholar
Digital Library
- Mahout. http://mahout.apache.org/.Google Scholar
- D. Mangot. Ec2 variability: The numbers revealed. http://goo.gl/NAH2lm.Google Scholar
- J. Mars and L. Tang. Whare-map: heterogeneity in "homogeneous" warehouse-scale computers. In Proceedings of the International Symposium on Computer Architecture (ISCA). Tel-Aviv, Israel, 2013.Google Scholar
Digital Library
- J. Mars, L. Tang, R. Hundt, K. Skadron, and M. L. Soffa. Bubble-up: increasing utilization in modern warehouse scale computers via sensible co-locations. In Proceedings of the IEEE/ACM International Symposium on Microarchitecture (MICRO). Porto Alegre, Brazil, 2011.Google Scholar
Digital Library
- S. M. Martin, K. Flautner, T. Mudge, and D. Blaauw. Combined dynamic voltage scaling and adaptive body biasing for lower power microprocessors under dynamic workloads. In Proceedings of International Conference On Computer Aided Design (ICCAD). 2002.Google Scholar
- R. Nathuji, C. Isci, and E. Gorbatov. Exploiting platform heterogeneity for power efficient data centers. In Proceedings of the International Conference on Autonomic Computing (ICAC). Jacksonville, FL, 2007.Google Scholar
Digital Library
- R. Nathuji, A. Kansal, and A. Ghaffarkhah. Q-clouds: Managing performance interference effects for qos-aware clouds. In Proceedings of EuroSys France, 2010.Google Scholar
Digital Library
- H. Nguyen, Z. Shen, X. Gu, S. Subbiah, and J. Wilkes. Agile: Elastic distributed resource scaling for infrastructure-as-a-service. In Proceedings of the International Conference on Autonomic Computing (ICAC). 2013.Google Scholar
- D. Novakovic, N. Vasic, S. Novakovic, D. Kostic, and R. Bianchini. Deepdive: Transparently identifying and managing performance interference in virtualized environments. In Proceedings of the USENIX Annual Technical Conference (ATC). San Jose, CA, 2013.Google Scholar
Digital Library
- S. Ostermann, A. Iosup, N. Yigitbasi, R. Prodan, T. Fahringer, and D. Epema. A performance analysis of ec2 cloud computing services for scientific computing. In Lecture Notes on Cloud Computing. Volume 34, p.115--131, 2010.Google Scholar
Cross Ref
- Z. Ou, H. Zhuang, J. K. Nurminen, A. Yla-Jaaski, and P. Hui. Exploiting hardware heterogeneity within the same instance type of amazon ec2. In HotCloud. 2012.Google Scholar
- K. Ousterhout, P. Wendell, M. Zaharia, and I. Stoica. Sparrow: Distributed, low latency scheduling. In Proceedings of the Symposium on Operating Systems Principles (SOSP). Farminton, PA, 2013.Google Scholar
Digital Library
- M. Rehman and M. Sakr. Initial findings for provisioning variation in cloud computing. In Proceedings of CloudCom. Indianapolis, IN, 2010.Google Scholar
Digital Library
- C. Reiss, A. Tumanov, G. Ganger, R. Katz, and M. Kozych. Heterogeneity and dynamicity of clouds at scale: Google trace analysis. In Proceedings of ACM Symposium on Cloud Computing (SOCC). 2012.Google Scholar
Digital Library
- Rightscale. https://aws.amazon.com/solution-providers/isv/rightscale.Google Scholar
- J. Schad, J. Dittrich, and J.-A. Quiané-Ruiz. Runtime measurements in the cloud: Observing, analyzing, and reducing variance. Proceedings VLDB Endow., 3(1--2):460--471, Sept. 2010.Google Scholar
Digital Library
- M. Schwarzkopf, A. Konwinski, M. Abd-El-Malek, and J. Wilkes. Omega: flexible, scalable schedulers for large compute clusters. In Proceedings of EuroSys. 2013.Google Scholar
Digital Library
- Z. Shen, S. Subbiah, X. Gu, and J. Wilkes. Cloudscale: elastic resource scaling for multi-tenant cloud systems. In Proceedings of ACM Symposium on Cloud Computing (SOCC). Cascais, Portugal, 2011.Google Scholar
Digital Library
- T. Simunic and S. Boyd. Managing power consumption in networks on chips. In Proceedings of Design Automation Conference (DAC). Paris, France, 2002.Google Scholar
Cross Ref
- S. Swanson, A. Putnam, M. Mercaldi, K. Michelson, A. Petersen, A. Schwerin, M. Oskin, and S. J. Eggers. Area-performance trade-offs in tiled dataflow architectures. In Proceedings of ACM SIGARCH Computer Architecture News, v.34 n.2, p.314--326, May 2006.Google Scholar
Digital Library
- K. Therdsteerasukdi, G. Byun, J. Cong, F. Chang, and G. Reinman. Utilizing rf-i and intelligent scheduling for better throughput/watt in a mobile gpu memory system. In Transactions on Architecture and Code Optimization (TACO) 8(4). 2012.Google Scholar
- Torque resource manager. http://www.adaptivecomputing.com/products/open-source/torque/.Google Scholar
- N. Vasić, D. Novaković, S. Miucin, D. Kostić, and R. Bianchini. Dejavu: accelerating resource allocation in virtualized environments. In Proceedings of the Seventeenth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS). 2012.Google Scholar
Digital Library
- G. Wang and T. S. E. Ng. The impact of virtualization on network performance of amazon ec2 data center. In Proceeedings of IEEE International Conference on Computer Communications (INFOCOM). San Diego, CA, 2010.Google Scholar
Cross Ref
- Windows azure. http://www.windowsazure.com/.Google Scholar
- H. Yang, A. Breslow, J. Mars, and L. Tang. Bubble-flux: precise online qos management for increased utilization in warehouse scale computers. In Proceedings of the International Symposium on Computer Architecture (ISCA). 2013.Google Scholar
Digital Library
- M. Zaharia, M. Chowdhury, T. Das, A. Dave, J. Ma, M. McCauley, M. Franklin, S. Shenker, and I. Stoica. Spark: Cluster computing with working sets. In Proceedings of the USENIX Sumposium on Networked Systems Design and Implementation (NSDI). San Jose, CA, 2012.Google Scholar
- S. M. Zahed and B. C. Lee. Ref: Resource elasticity fairness with sharing incentives for multiprocessors. In Proceedings of the Nineteenth International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS). Salt Lake City, UT, 2014.Google Scholar
Digital Library
- J. Y. Zhang, P. Wu, J. Zhu, H. Hu, and F. Bonomi. Privacy-preserved mobile sensing through hybrid cloud trust framework. In Proceedings of IEEE International Conference on Cloud Computing (CLOUD). 2013.Google Scholar
Digital Library
- X. Zhang, E. Tune, R. Hagmann, R. Jnagal, V. Gokhale, and J. Wilkes. Cpi2: Cpu performance isolation for shared compute clusters. In Proceedings of EuroSys. Prague, 2013.Google Scholar
Digital Library
- L. Zhao, R. Iyer, J. Moses, R. Illikkal, S. Makineni, and D. Newell. Exploring large-scale cmp architectures using manysim. In IEEE Micro, vol.27 n.4, July 2007.Google Scholar
Digital Library
Index Terms
HCloud: Resource-Efficient Provisioning in Shared Cloud Systems
Recommendations
HCloud: Resource-Efficient Provisioning in Shared Cloud Systems
ASPLOS '16: Proceedings of the Twenty-First International Conference on Architectural Support for Programming Languages and Operating SystemsCloud computing promises flexibility and high performance for users and cost efficiency for operators. To achieve this, cloud providers offer instances of different sizes, both as long-term reservations and short-term, on-demand allocations. ...
HCloud: Resource-Efficient Provisioning in Shared Cloud Systems
ASPLOS'16Cloud computing promises flexibility and high performance for users and cost efficiency for operators. To achieve this, cloud providers offer instances of different sizes, both as long-term reservations and short-term, on-demand allocations. ...
Modeling and Analysis of Cloud Computing Availability Based on Eucalyptus Platform for E-Government Data Center
IMIS '11: Proceedings of the 2011 Fifth International Conference on Innovative Mobile and Internet Services in Ubiquitous ComputingData center is required for the enterprise, cloud service providers such as Amazon Web Service and Microsoft Azure offer enterprises the ability to store their data remotely on rented hardware per demand. But designing cloud infrastructure in term for ...







Comments