Tian Guo
Upendra Sharma
Prashant Shenoy
Abstract
The high cost of provisioning resources to meet peak application demands has led to the widespread adoption of pay-as-you-go cloud computing services to handle workload fluctuations. Some enterprises with existing IT infrastructure employ a hybrid cloud model where the enterprise uses its own private resources for the majority of its computing, but then “bursts” into the cloud when local resources are insufficient. However, current commercial tools rely heavily on the system administrator’s knowledge to answer key questions such as when a cloud burst is needed and which applications must be moved to the cloud. In this article, we describe Seagull, a system designed to facilitate cloud bursting by determining which applications should be transitioned into the cloud and automating the movement process at the proper time. Seagull optimizes the bursting of applications using an optimization algorithm as well as a more efficient but approximate greedy heuristic. Seagull also optimizes the overhead of deploying applications into the cloud using an intelligent precopying mechanism that proactively replicates virtualized applications, lowering the bursting time from hours to minutes. Our evaluation shows over 100% improvement compared to naïve solutions but produces more expensive solutions compared to ILP. However, the scalability of our greedy algorithm is dramatically better as the number of VMs increase. Our evaluation illustrates scenarios where our prototype can reduce cloud costs by more than 45% when bursting to the cloud, and that the incremental cost added by precopying applications is offset by a burst time reduction of nearly 95%.
- M. Armbrust, A. Fox, R. Griffith, A. D. Joseph, R. H. Katz, A. Konwinski, G. Lee, D. Patterson, A. Rabkin, I. Stoica, and M. Zaharia. 2009. Above the couds: A Berkeley view of cloud computing. Technical Report UCB/EECS-2009-28. EECS Department, University of California, Berkeley. http://www.eecs.berkeley.edu/Pubs/TechRpts/2009/EECS-2009-28.html.Google Scholar
- AWSECO 2013. AWS economics center. http://aws.amazon.com/economics/.Google Scholar
- H. Ballani, P. Costa, T. Karagiannis, and A. Rowstron. 2011. Towards predictable datacenter networks. In Proceedings of the ACM SIGCOMM Conference (SIGCOMM’11). ACM, 242--253. Google ScholarDigital Library
- T. Bicer, D. Chiu, and G. Agrawal. 2011. A framework for data-intensive computing with cloud bursting. In Proceedings of the IEEE International Conference on Cluster Computing (CLUSTER). 169--177. DOI:http://dx.doi.org/10.1109/CLUSTER.2011.21. Google ScholarDigital Library
- R. Bradford, E. Kotsovinos, A. Feldmann, and H. Schiöberg. 2007. Live wide-area migration of virtual machines including local persistent state. In Proceedings of VEE. ACM, 169--179. DOI:http://dx.doi.org/10.1145/1254810.1254834. Google ScholarDigital Library
- R. Buyya, R. Ranjan, and R. N. Calheiros. 2010. InterCloud: Utility-oriented federation of cloud computing environments for scaling of application services. In Proceedings of the International Conference on Algorithms and Architectures for Parallel Processing. Google ScholarDigital Library
- E. Cecchet, V. Udayabhanu, T. Wood, and P. Shenoy. 2011. BenchLab: An open testbed for realistic benchmarking of web applications. In Proceedings of the 2nd USENIX Conference on Web Application Development (WebApps). Google ScholarDigital Library
- C. Clark, K. Fraser, S. Hand, J. G. Hansen, E. Jul, C. Limpach, I. Pratt, and A. Warfield. 2005. Live migration of virtual machines. In Proceedings of NSDI. Google ScholarDigital Library
- Cloudbursting 2008. Cloudbursting - Hybrid Application Hosting. http://aws.typepad.com/aws/2008/08/cloudbursting-.html. (Last accessed 8/08).Google Scholar
- E. G. Coffman, Jr., M. R. Garey, and D. S. Johnson. 1997. Approximation algorithms for bin packing: A survey. In Approximation Algorithms for NP-Hard Problems. Google ScholarDigital Library
- K. A. Dowsland and W. B. Dowsland. 1992. Packing problems. Europ. J. Oper. Res. 56, 1, 2--14.Google ScholarCross Ref
- A. Gulati, G. Shanmuganathan, I. Ahmad, C. Waldspurger, and M. Uysal. 2011. Pesto: Online storage performance management in virtualized datacenters. In Proceedings of SOCC (SOCC’11). ACM, Article 19. DOI:http://dx.doi.org/10.1145/2038916.2038935. Google ScholarDigital Library
- C. Guo, G. Lu, H. J. Wang, S. Yang, C. Kong, P. Sun, W. Wu, and Y. Zhang. 2010. SecondNet: A data center network virtualization architecture with bandwidth guarantees. In Proceedings of the 6th International Conference (Co-NEXT’10). ACM, Article 15. Google ScholarDigital Library
- J. Hellerstein, F. Zhang, and P. Shahabuddin. 1999. An approach to predictive detection for service management. In Proceedings of the IEEE International Conference on Systems and Network Management.Google Scholar
- S. Kailasam, N. Gnanasambandam, J. Dharanipragada, and N. Sharma. 2010. Optimizing service level agreements for autonomic cloud bursting schedulers. In Proceedings of the 39th International Conference on Parallel Processing Workshops (ICPPW ’10). IEEE Computer Society, 285--294. Google ScholarDigital Library
- H. Kim, M. Parashar, D. J. Foran, and L. Yang. 2009. Investigating the use of autonomic cloudbursts for high-throughput medical image registration. In Proceedings of GRID. IEEE, 34--41. Google ScholarDigital Library
- G. Lee, N. Tolia, P. Ranganathan, and R. H. Katz. 2010. Topology-aware resource allocation for data-intensive workloads. In Proceedings of the 1st ACM Asia-Pacific Workshop on Systems (APSys’10). ACM, 1--6. Google ScholarDigital Library
- A. Mashtizadeh, E. Celebi, T. Garfinkel, and M. Cai. 2011. The design and evolution of live storage migration in VMware ESX. In Proceedings of USENIX ATC. 14--14. http://dl.acm.org/citation.cfm?id=2002181.2002195. Google ScholarDigital Library
- G. Mateescu, W. Gentzsch, and C. J. Ribbens. 2011. Hybrid computing-where HPC meets grid and cloud computing. Future Gener. Comput. Syst. 27, 5, 440--453. DOI:http://dx.doi.org/10.1016/j.future.2010.11.003. Google ScholarDigital Library
- M. Mishra and A. Sahoo. 2011. On theory of VM placement: Anomalies in existing methodologies and their mitigation using a novel vector based approach. In Proceedings of the IEEE International Conference on Cloud Computing (CLOUD), 275--282. DOI:http://dx.doi.org/10.1109/CLOUD.2011.38. Google ScholarDigital Library
- K. Nagin, D. Hadas, Z. Dubitzky, A. Glikson, I. Loy, B. Rochwerger, and L. Schour. 2011. Inter-cloud mobility of virtual machines. In Proceedings of the Annual International Conference on Systems and Storage (SYSTOR’11). ACM, Article 3. DOI:http://dx.doi.org/10.1145/1987816.1987820. Google ScholarDigital Library
- M. Nelson, B.-H. Lim, and G. Hutchins. 2005. Fast transparent migration for virtual machines. In Proceedings of ATEC ’05: USENIX ATC. USENIX Association, Berkeley, CA, 25. Google ScholarDigital Library
- ObjectWeb. The ObjectWeb TPC-W implementation. Website. http://jmob.objectweb.org/tpcw.html.Google Scholar
- OpenNebula 2012. The Open Source Toolkit for Data Center Virtualization. (2012). http://www.opennebula.org/.Google Scholar
- Openstack 2012. openstack: Cloud Software. http://www.openstack.org/.Google Scholar
- A. Rai, R. Bhagwan, and S. Guha. 2012. Generalized resource allocation for the cloud. In Proceedings of the 3rd ACM Symposium on Cloud Computing (SoCC’12). ACM, Article 15, 12 pages. Google ScholarDigital Library
- S. Ranjan, J. Rolia, H. Fu, and E. Knightly. 2002. QoS-driven server migration for Internet data centers. In Proceedings of IWQoS. 3--12.Google Scholar
- B. Rochwerger, D. Breitgand, A. Epstein, et al. 2011. Reservoir - When one cloud is not enough. Computer 44, 44--51. DOI:http://dx.doi.org/10.1109/MC.2011.64. Google ScholarDigital Library
- U. Sharma, P. Shenoy, S. Sahu, and A. Shaikh. 2011. Kingfisher: Cost-aware elasticity in the cloud. In Proceedings of INFOCOM 2011. 206--210. DOI:http://dx.doi.org/10.1109/INFCOM.2011.5935016.Google ScholarCross Ref
- Z. Shen, S. Subbiah, X. Gu, and J. Wilkes. 2011. CloudScale: Elastic resource scaling for multi-tenant cloud systems. In Proceedings of (SOCC’11). ACM, Article 5, 14 pages. DOI:http://dx.doi.org/10.1145/2038916.2038921. Google ScholarDigital Library
- P. Shivam, A. Iamnitchi, A. R. Yumerefendi, and J. S. Chase. 2005. Model-driven placement of compute tasks and data in a networked utility. In Proceedings of ICAC. DOI:http://dx.doi.org/10.1109/ICAC.2005.41. Google ScholarDigital Library
- W. Sobel, S. Subramanyam, A. Sucharitakul, J. Nguyen, H. Wong, S. Patil, A. Fox, and D. Patterson. 2008. Cloudstone: Multi-platform, multi-language benchmark and measurement tools for web 2.0. In Proceedings of Cloud Computing and its Applications.Google Scholar
- B. Sotomayor, R. S. Montero, I. M. Llorente, and I. Foster. 2009. Virtual infrastructure management in private and hybrid clouds. Internet Comput. 13, 5, 14--22. Google ScholarDigital Library
- Terremark 2012. Study: USA.gov Achieves cloud bursting efficiency using terremark enterprise cloud. http://terremark.com.Google Scholar
- B. Urgaonkar, G. Pacifici, P. Shenoy, M. Spreitzer, and A. Tantawi. 2005. An analytical model for multi-tier internet services and its applications. In Proceedings of the ACM Sigmetrics Conference. Google ScholarDigital Library
- B. Urgaonkar, P. Shenoy, A. Chandra, P. Goyal, and T. Wood. 2008. Agile dynamic provisioning of multi-tier Internet applications. ACM Trans. Auton. Adapt. Syst. 3, Article 1. Google ScholarDigital Library
- VDataCenter 2012. VMware: Public & hybrid cloud computing. http://www.vmware.com/solutions/cloud-computing/public-cloud/products.html.Google Scholar
- VMotion 2009. Virtual machine mobility with VMware VMotion and Cisco data center Interconnect technologies. http://www.cisco.com/en/US/solutions/collateral/ns340/ns517/ns224/ns836/white_paper_c11-557822.pdf.Google Scholar
- VMware DRS. Resource management with VMware DRS. http://www.vmware.com/pdf/vmware_drs_wp.pdf.Google Scholar
- D. Williams, H. Jamjoom, and H. Weatherspoon. 2012. The Xen-Blanket: Virtualize once, run everywhere. In Proceedings of the 7th ACM European Conference on Computer Systems (EuroSys’12). 113--126. Google ScholarDigital Library
- T. Wood, K. K. Ramakrishnan, P. Shenoy, and J. Van der Merwe. 2011. CloudNet: Dynamic pooling of cloud resources by live WAN migration of virtual machines. In Proceedings of VEE. 121--132. DOI:http://dx.doi.org/10.1145/1952682.1952699. Google ScholarDigital Library
- T. Wood, P. Shenoy, A. Venkataramani, and M. Yousif. 2009. Sandpiper: Black-box and gray-box resource management for virtual machines. Comput. Netw. The Int. J. Comput. Telecom. Netw. 53, 17. http://portal.acm.org/citation.cfm?id=1663647.1663710. Google ScholarDigital Library
- Z. Xiao, W. Song, and Q. Chen. 2013. Dynamic resource allocation using virtual machines for cloud computing environment. IEEE Trans. Parall. Distrib. Syst. 24, 6, 1107--1117. DOI:http://dx.doi.org/10.1109/TPDS.2012.283. Google ScholarDigital Library
- Y. Zhang, V. Paxson, and S. Shenkar. 2000. The stationarity of Internet path properties: Routing, loss, and throughput. Technical Report. AT&T Center for Internet Research at ICSI, http://www.aciri.org/.Google Scholar
- J. Zheng, T. E. Ng, and K. Sripanidkulchai. 2011. Workload-aware live storage migration for clouds. In Proceeding of VEE (VEE’11). ACM, 133--144. DOI:http://dx.doi.org/10.1145/1952682.1952700. Google ScholarDigital Library
Index Terms
- Cost-Aware Cloud Bursting for Enterprise Applications
Recommendations
VirtualWire: system support for live migrating virtual networks across clouds
VTDC '13: Proceedings of the 7th international workshop on Virtualization technologies in distributed computingDespite significant advances in enabling live virtual machine (VM) migration within a virtualized--cloud--infrastructure, cross-cloud live migration remains an ad hoc, complex process. To create a network environment in which live migration can occur, ...
A cost model for hybrid clouds
GECON'11: Proceedings of the 8th international conference on Economics of Grids, Clouds, Systems, and ServicesCloud computing aims at allowing customers to utilize computational resources and software hosted by service providers. Thus, it shifts the complex and tedious resource and software management tasks typically done by customers to the service providers. ...
The Aneka platform and QoS-driven resource provisioning for elastic applications on hybrid Clouds
Cloud computing alters the way traditional software systems are built and run by introducing a utility-based model for delivering IT infrastructure, platforms, applications, and services. The consolidation of this new paradigm in both enterprises and ...
Reviews
Jill Gemmill
Seagull, a system to optimally manage cloud bursting, is described in this paper; optimization is with respect to the number and size of virtual machine (VM) images and total data transfer. A performance evaluation of Seagull is also provided. The authors address an important research topic in cloud computing, exploring how to manage applications with varying resource requirements in an automated manner that also minimizes costs in terms of both dollars and application performance. Seagull is designed for use by enterprise data centers that require bursting into cloud resources for occasional excess capacity demands. The paper offers a valuable contribution to a difficult and reality-based problem. The authors use some insight in designing their approach: (1) When an application that is large (lots of data, many VMs) requires more resources than are available locally, it may be better to move some other, smaller apps into the cloud, freeing up local resources. (2) Since the biggest delay in restarting an app that has been moved comes from having to move the data (app data and virtual machine images), occasional precopying of the smaller apps with a high likelihood of being moved into the cloud will significantly reduce the delay caused by the need to copy. Seagull addresses the classic bin-packing problem, which is known to be NP-hard. An optimal integer linear program (ILP) formulation is introduced and an algorithm for precopying is presented; in addition, an algorithm using a greedy heuristic is provided for approximating an optimal answer for large-scale problems. The heuristic uses sorting to improve results. The authors also present a prototyped Seagull using a Xen-based local data center and Amazon EC2 for cloud bursting. Finally, a detailed experimental evaluation for Seagull is presented, using three examples of different web applications. The applications each have a MySQL database backend; one is a two-tier Java application implemented as an Apache Tomcat servlet, one uses a PHP application, and one uses an Ajax application and supports a memcached tier that can be horizontally scaled. The authors testbedprototype choices (for example, choice of apps, parameterssizes selected) are well reasoned and explained. The design is flexible in that different algorithmsweighting representing cost and excess demand can be substituted. The experimental protocol is thorough and well reasoned. Figure 12 shows Seagulls scalability to 800 hosts with proportional increases in numbers of the three applications. Real data centers may be running hundreds (or thousands) of applications, as well as thousands of hosts. It would be interesting to think through how Seagulls performance would be impacted by this much longer list of applications to consider as part of the computation time. Overall, this was an excellent and very thorough paper. Online Computing Reviews Service
Access critical reviews of Computing literature here
Become a reviewer for Computing Reviews.
Comments