ABSTRACT
Modern resource management frameworks guarantee low tail latency for long-running services using the resource over-provisioning method, resulting in serious waste of resource and increasing the service costs greatly. To reduce the over-provisioning cost, we present EFRA, an elastic and fine-grained resource allocator that enables much more efficient resource provisioning while guaranteeing the tail latency Service Level Objective (SLO). EFRA achieves this through the cooperation of three key components running on a containerized platform: The period detector identifies the period features of the workload through a convolution-based time series analysis. The resource reservation component estimates the just-right amount of resources based on the period analysis through a top-K based collaborative filtering approach. The online reprovisioning component dynamically adjusts the resources for further enforcing the tail latency SLO. Testbed experiments show that EFRA is able to increase the average resource utilization to 43%, and save up to 66% resources while guaranteeing the same tail latency objective.
- 2017. Alibaba trace. (2017). https://github.com/alibaba/clusterdata.Google Scholar
- 2017. Docker platform. (2017). https://www.docker.com/docker-engine.Google Scholar
- 2017. The Internet Traffic Archive. (2017). http://ita.ee.lbl.gov/html/traces.html.Google Scholar
- Omer Adam, Young Choon Lee, et al. 2017. Stochastic Resource Provisioning for Containerized Multi-Tier Web Services in Clouds. IEEE Transactions on Parallel and Distributed Systems 28, 7 (2017), 2060--2073.Google ScholarDigital Library
- Omer Y. Adam, Young Choon Lee, et al. 2016. Constructing Performance-Predictable Clusters with Performance-Varying Resources of Clouds. IEEE Trans. Comput. 65, 9 (Sept. 2016), 2709--2724. Google ScholarDigital Library
- M. J. Atallah, F. Chyzak, et al. 2001. A Randomized Algorithm for Approximate String Matching. Algorithmica 29, 3 (2001), 468--486. Google ScholarDigital Library
- Berk Atikoglu, Yuehai Xu, et al. 2012. Workload Analysis of a Large-scale Key-value Store. In SIGMETRICS '12. ACM, New York, NY, USA, 53--64. Google ScholarDigital Library
- Christian Bliek1ú, Pierre Bonami, et al. 2014. Solving mixed-integer quadratic programming problems with IBM-CPLEX: a progress report. In Proceedings of the Twenty-Sixth RAMP Symposium. 171--180.Google Scholar
- John S. Breese, David Heckerman, et al. 1998. Empirical Analysis of Predictive Algorithms for Collaborative Filtering. In UAI'98. Morgan Kaufmann Publishers Inc., San Francisco, CA, USA, 43--52. Google ScholarDigital Library
- Ludmila Cherkasova. 2011. Performance Modeling in Mapreduce Environments: Challenges and Opportunities. In ICPE '11. ACM, New York, NY, USA, 5--6. Google ScholarDigital Library
- Eli Cortez, Anand Bonde, et al. 2017. Resource Central: Understanding and Predicting Workloads for Improved Resource Management in Large Cloud Platforms. In SOSP '17. ACM, New York, NY, USA, 153--167. Google ScholarDigital Library
- Carlo Curino, Djellel E. Difallah, et al. 2014. Reservation-based Scheduling: If You'Re Late Don'T Blame Us!. In SOCC '14. ACM, New York, NY, USA, Article 2, 14 pages. Google ScholarDigital Library
- Christina Delimitrou and Christos Kozyrakis. 2014. Quasar: Resource-efficient and QoS-aware Cluster Management. In ASPLOS '14. ACM, New York, NY, USA, 127--144. Google ScholarDigital Library
- Christina Delimitrou and Christos Kozyrakis. 2016. HCloud: Resource-Efficient Provisioning in Shared Cloud Systems. In ASPLOS '16. ACM, New York, NY, USA, 473--488. Google ScholarDigital Library
- Mohamed G. Elfeky, Walid G. Aref, et al. 2005. Periodicity Detection in Time Series Databases. IEEE Trans. on Knowl. and Data Eng. 17, 7 (July 2005), 875--887. Google ScholarDigital Library
- Anshul Gandhi, Yuan Chen, et al. 2011. Minimizing data center SLA violations and power consumption via hybrid resource provisioning. In Green Computing Conference and Workshops. 1--8. Google ScholarDigital Library
- D Gmach, J Rolia, L Cherkasova, et al. 2009. An integrated approach to resource pool management: Policies, efficiency and quality metrics. In IEEE International Conference on Dependable Systems and Networks with Ftcs and DCC. 326--335.Google Scholar
- Zhenhuan Gong, Xiaohui Gu, et al. 2010. PRESS: PRedictive Elastic ReSource Scaling for cloud systems. In International Conference on Network and Service Management. 9--16.Google Scholar
- Benjamin Hindman, Andy Konwinski, et al. 2011. Mesos: A Platform for Finegrained Resource Sharing in the Data Center. In NSDI'11. USENIX Association, Berkeley, CA, USA, 295--308. Google ScholarDigital Library
- Sangeetha Abdu Jyothi, Carlo Curino, et al. 2016. Morpheus: Towards Automated SLOs for Enterprise Clusters. In OSDI'16. USENIX Association, Berkeley, CA, USA, 117--134. Google ScholarDigital Library
- Palden Lama and Xiaobo Zhou. 2012. AROMA: Automated Resource Allocation and Configuration of Mapreduce Environment in the Cloud. In ICAC '12. ACM, New York, NY, USA, 63--72. Google ScholarDigital Library
- Richard J Larsen and Morris L Marx. 1981. An introduction to mathematical statistics and its applications. Prentice-Hall. 2061--2071 pages.Google Scholar
- Jessica Lin, Eamonn Keogh, et al. 2003. A Symbolic Representation of Time Series, with Implications for Streaming Algorithms. In DMKD '03. ACM, New York, NY, USA, 2--11. Google ScholarDigital Library
- Haikun Liu and Bingsheng He. 2014. Reciprocal Resource Fairness: Towards Cooperative Multiple-resource Fair Sharing in IaaS Clouds. In SC '14. IEEE Press, Piscataway, NJ, USA, 970--981. Google ScholarDigital Library
- David Lo, Liqun Cheng, et al. 2016. Improving Resource Efficiency at Scale with Heracles. ACM Trans. Comput. Syst. 34, 2, Article 6 (May 2016), 33 pages. Google ScholarDigital Library
- Richard Mccreadie, Ian Soboroff, et al. 2012. On building a reusable Twitter corpus. In International ACM SIGIR conference on research and development in Information Retrieval. 1113--1114. Google ScholarDigital Library
- Malte Schwarzkopf, Andy Konwinski, et al. 2013. Omega: Flexible, Scalable Schedulers for Large Compute Clusters. In EuroSys '13. ACM, New York, NY, USA, 351--364. Google ScholarDigital Library
- Zhiming Shen, Sethuraman Subbiah, et al. 2011. CloudScale: Elastic Resource Scaling for Multi-tenant Cloud Systems. In SOCC '11. ACM, New York, NY, USA, Article 5, 14 pages. Google ScholarDigital Library
- Sethuraman Subbiah, John Wilkes, et al. 2013. AGILE: Elastic Distributed Resource Scaling for Infrastructure-as-a-Service. In International Conference on Autonomic Computing.Google Scholar
- Alexey Tumanov, Timothy Zhu, et al. 2016. TetriSched: Global Rescheduling with Adaptive Plan-ahead in Dynamic Heterogeneous Clusters. In EuroSys '16. ACM, New York, NY, USA, Article 35, 16 pages. Google ScholarDigital Library
- Guido Urdaneta, Guillaume Pierre, et al. 2009. Wikipedia Workload Analysis for Decentralized Hosting. Comput. Netw. 53, 11 (July 2009), 1830--1845. Google ScholarDigital Library
- Vinod Kumar Vavilapalli, Arun C. Murthy, et al. 2013. Apache Hadoop YARN: Yet Another Resource Negotiator. In SOCC '13. ACM, New York, NY, USA, Article 5, 16 pages. Google ScholarDigital Library
- Abhishek Verma, Ludmila Cherkasova, et al. 2011. ARIA: Automatic Resource Inference and Allocation for Mapreduce Environments. In ICAC '11. ACM, New York, NY, USA, 235--244. Google ScholarDigital Library
- Abhishek Verma, Luis Pedrosa, et al. 2015. Large-scale Cluster Management at Google with Borg. In EuroSys '15. ACM, New York, NY, USA, Article 18, 17 pages. Google ScholarDigital Library
- Yunqi Zhang, George Prekas, et al. 2016. History-based Harvesting of Spare Cycles and Storage in Large-scale Datacenters. In OSDI '16. USENIX Association, Berkeley, CA, USA, 755--770. Google ScholarDigital Library
- Timothy Zhu, Daniel S. Berger, et al. 2016. SNC-Meister: Admitting More Tenants with Tail Latency SLOs. In SoCC '16. ACM, New York, NY, USA, 374--387. Google ScholarDigital Library
Recommendations
Cloud resource provisioning: survey, status and future research directions
Cloud resource provisioning is a challenging job that may be compromised due to unavailability of the expected resources. Quality of Service (QoS) requirements of workloads derives the provisioning of appropriate resources to cloud workloads. Discovery ...
Resource provisioning and scheduling in clouds: QoS perspective
Resource provisioning of appropriate resources to cloud workloads depends on the quality of service (QoS) requirements of cloud applications and is a challenging task. In cloud environment, heterogeneity, uncertainty and dispersion of resources ...
User's priority focused resource provisioning over cloud computing infrastructure
Resource provisioning is the process of activating a bundle of allocated quantity of resources to bear the user requests. The scheduling algorithm plays a vital role in effective utilisation of resources, though resource allocation fails to achieve user ...
Comments