CAREER: Building Resilient Internet Services with Learning and Control (NSF CNS-0844983, 09/2009-08/2014)

Project description and goals

Due to the dynamic nature and unprecedented scale of the Internet, Internet services pose challenges including scalability, reliability, and availability to underlying networked systems. This CAREER project, directed by Prof. Xiaobo Zhou, concentrates on building Internet services that are resilient to those challenges with machine learning and control techniques. Internet services build upon cluster-based computer systems that keep growing in scale and complexity. Such systems become so complicated that it is even a big challenge to get a good understanding of the entire system dynamic behaviors. The investigators take an analytical and organized approach to design an autonomous software infrastructure on networked systems for building resilient Internet services. The project builds empirical models using statistical learning to help overcome the challenges of scale and complexity in networked systems. It designs coordinated admission control and capacity planning algorithms with end-to-end quality-of-service on multi-tier clusters. Model-independent control techniques are used with empirical models to allocate resources and to dynamically reconfigure the system for performance optimization needs. It develops performance differentiation, isolation, and self-adaptive reconfiguration capabilities for enhancing system reliability and availability. It broadens the research impact by developing a testbed in a data center lab to demonstrate the orchestration of designed techniques for automated arrangement, coordination, and management of complex computer systems, middleware, and services.

The research project is exectued in a cutting-edge lab located in the new science and engineering building. The server room is furnished with cutting-edge HP data center blade facility that has three racks of HP ProLiant BL460C G6 blade server modules and a 40 TB HP EVA storage area network with 10 Gbps Ethernet and 8 Gbps Fibre/iSCSI dual channels. It has three APC InRow RP Air-Cooled and UPS equipments for maximum 40 kWs in the n+1 redundancy design.

This five-year project was kicked off in September 2009.

You can view this page in Romanian.

You can view this page in Portuguese for https://www.homeyou.com/~edu/

Participants

• Dr. Xiaobo Zhou, The Principal Investigator
• Palden Lama, PhD student, 2008 - 2013 (now assistant professor in UTSA)
• Sireesha Muppala, PhD student, 2007 - 2013 (now with Amazon)
• Dennis Ippoliti, PhD student, 2007 - 2013 (now with Microsoft)
• Yanfei Guo, PhD student, 2010 - 2015 (now with Argonne National Lab)
• Dazhao Cheng, PhD student, 2011 - 2016 (now assistant professor in UNCC)
• Matt Nesterenko, Undergraduate student, 2011 - 2012
• Theodore Doss, Undergraduate student, 2010 - 2011

Project-sponsored Publications (Download Bibtex)

• "Towards Energy Efficiency in Heterogeneous Hadoop Clusters by Adaptive Task Assignment", Dazhao Cheng, Palden Lama, Changjun Jiang, and Xiaobo Zhou, Proc. of the 35th IEEE ICDCS (acceptance rate 12.8%), 10 pages, June/July 2015.
• "Heterogeneity-aware Workload Placement and Migration in Distributed Sustainable Datacenters", Dazhao Cheng, Changjun Jiang, and Xiaobo Zhou, Proc. of the 28th IEEE IPDPS (acceptance rate 21%), 10 pages, May 2014.
• "Optimizing Virtual Machine Scheduling in NUMA Multicore Systems", Jia Rao and Xiaobo Zhou, In Proc. of the 19th ACM PPoPP (acceptance rate 16%), 10 pages, Feb 2014.
• "Multi-tier Service Differentiation by Coordinated Learning-based Resource Provisioning and Admission Control", accepted by the Journal of Paralleland Distributed Computing, Jan 2014.
• "Coordinated Power and Performance Guarantee with Fuzzy MIMO Control in Virtualized Server Clusters", Palden Lama and Xiaobo Zhou, IEEE Transactions on Computers, 14 pages, accepted, July 2013.
• "Autonomic Provisioning with Self-Adaptive Neural Fuzzy Control for Percentile-based Delay Guarantee", Palden Lama and Xiaobo Zhou, ACM Transactions on Autonomous and Adaptive Systems, 8(2):1-31, July 2013.
• "iShuffle: Improving Hadoop Performance with Shuffle-on-Write", Yanfei Guo, Jia Rao, and Xiaobo Zhou, won the Best Paper Award of the 10th USENIX ICAC, 11 pages, San Jose, June 2013.
• "V-Cache: Towards Flexible Resource Provisioning for Multi-tier Applications in IaaS Clouds", Yanfei Guo, Palden Lama, Jia Rao, and Xiaobo Zhou, In Proc. of the 27th IEEE IPDPS, 12 pages, Boston, May 2013 (acceptance rate 21%).
• "Self-tuning Batching with DVFS for Improving Performance and Energy Efficiency in Servers", Dazhao Cheng, Yanfei Guo, and Xiaobo Zhou, In Proc. of the 21th IEEE MASCOTS (acceptance rate 27%), 10 pages, San Francisco, August 2013.
• "Power-Aware Dynamic Placement and Migration in Virtualized GPU Environments", Palden Lama, Yan Li, Ashwin Aji, Pavan Balaji, James Dinan, Shucai Xiao, Yunquan Zhang, Wwchun Feng, Rajeev Thakur, Xiaobo Zhou, In Proc. of the 33rd IEEE ICDCS, 10 pages, Philadephia, July 2013 (acceptance rate 13%).
• "Autonomic Performance and Power Control for Co-located Web Applications on Virtualized Servers", Palden Lama, Yanfei Guo, and Xiaobo Zhou, In Proc. of the 21st ACM/IEEE IWQoS (acceptance rate 28%), 10 pages, Montreal, June 2013.
• "Optimizing Virtual Machine Scheduling in NUMA Multicore Systems", Jia Rao, Kun Wang, Xiaobo Zhou, and Cheng-Zhong Xu, Proc. of the 19th IEEE HPCA (accpeptance rate 20%), 12 pages, Feb 2013. The paper is one of four Best Paper Award candidates (out of 249 submissions).
• Automated and Agile Server Parameter Tuning by Coordinated Learning and Control", Yanfei Guo, Palden Lama, Changjun Jiang, and Xiaobo Zhou, IEEE Transactions on Parallel and Distributed Systems, accepted, April 2013.
• "A-GHSOM: An Adaptive Growing Hierarchical Self Organizing Map for Network Anomaly Detection", Dennis Ippoliti and Xiaobo Zhou, Journal of Parallel and Distributed Computing, Elsevier, Vol. 72, No. 12, pages 1576-1590, December 2012.
• "NINPIN: Non-Invasive and Energy Efficient Performance Isolation in Virtualized Servers", Palden Lama and Xiaobo Zhou, 12 pages, In Proc. of the 42nd IEEE/IFIP DSN, June 2012 (acceptance rate 17%).
• "AROMA: Automated Resource Allocation and Configuration of MapReduce Environment in the Cloud", Palden Lama and Xiaobo Zhou, 10 pages, In Proc. of the 9th ACM ICAC, September 2012 (acceptance rate 22%).
• "Automated and Agile Server Parameter Tuning with Learning and Control", Yanfei Guo, Palden Lama and Xiaobo Zhou, 12 pages, In Proc. of the 26th IEEE IPDPS, May 2012 (acceptance rate 21%).
• "Coordinated VM Resizing and Server Tuning: Throughput, Power Efficiency and Scalability", Yanfei Guo and Xiaobo Zhou, In Proc. of the 20th IEEE MASCOTS (accpeptance rate 36%).
• "Multi-tier Service Differentiation: Coordinated Resource Provisioning and Admission Control", Sireesha Muppala, Xiaobo Zhou and Guihai Chen, In Proc. of the 18th IEEE ICPADS (accpeptance rate 30%), December 2012.
• "A Self-Tuning Self-Optimizing Approach for Automated Network Anomaly Detection Systems", Dennis Ippoliti and Xiaobo Zhou, In Proc. of the the 9th ACM ICAC, September 2012.
• "Efficient Server Provisioning with Control for End-to-End Response Time Guarantee on Multi-tier Clusters", Palden Lama and Xiaobo Zhou, IEEE Transactions on Parallel and Distributed Systems, Vol. 23, No. 1, 2012; Download Link; Talk slides (IWQoS'09).
• "Regression Based Resource Provisioning for Session Slowdown Guarantees in Multi-tier Internet Servers", Sireesha Muppala, Xiaobo Zhou, Liqiang Zhang and Guihai Chen, Journal of Parallel and Distributed Computing, Elsevier, accepted in Nov 2011; Download Link.
• Multi-tier Service Differentiation: Coordinated Resource Provisioning and Admission Control, Sireesha Muppala, Xiaobo Zhou and Guihai Chen, Proc. of the 18th IEEE International Conference on Parallel and Distributd Systems (ICPADS), Singapore, December 2012 (acceptance rate 30%)
• "PERFUME: Power and Performance Guarantee with Fuzzy MIMO Control in Virtualized Servers", Palden Lama and Xiaobo Zhou, Proc. of the 19th ACM/IEEE International Workshop on Quality-of-Service (IWQoS), pages 1-9, June 2011 (acceptance rate 28%); Download Link; Talk slides.
• "aMOSS: Automated Multi-Objective Server Provisioning with Stress-Strain Curving", Palden and Xiaobo Zhou, Proc. of the 40th IEEE International Conference on Parallel Processing (ICPP), pages 145-154, Sep 2011 (acceptance rate 21%). Download Link; Talk slides.
• "Enhanced Statistics-based Rate Aadapation for 802.11 in wireless networks", Liqiang Zhang, Yujen Cheng, and Xiaobo Zhou, Journal of Network and Computer Applications, Elsevier 34(5), Sep 2011; Download Link.
• "Coordinated Session-based Admission Control with Statistical Learning for Multi-tier Internet Applications", Sireesha Muppala and Xiaobo Zhou, Journal of Network and Computer Applications, Elsevier, Vol. 34, No. 1, pp. 20-29, Jan 2011; Download Link.
• "Autonomic Provisioning with Self-Adaptive Neural Fuzzy Control for End-to-end Delay Guarantee", Palden Lama and Xiaobo Zhou, Proceedings of the 18th IEEE/ACM International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunication Systems (MASCOTS 2010), pages 151-160, Miami, August 2010 (accepted as a long extended paper with the selection rate 16%); Download Link; Talk slides.
• "Regression Based Multi-tier Resource Provisioning for Session Slowdown Guarantees", Sireesha Muppala, Xiaobo Zhou and Liqiang Zhang, Proc. of the 29th IEEE International Performance Computing and Communications Conference (IPCCC), Albuquerque, December 2010 (acceptance rate 28%); Download Link; Talk slides.
• "An Adaptive Growing Hierarchical Self Organizing Map for Network Intrusion Detection", Dennis Ippoliti and Xiaobo Zhou, Proceedings of the 19th IEEE International Conference on Computer Communications and Networks (ICCCN 2010), Zurich, Swizerland, August 2010 (acceptance rate 30%); Download Link; Talk slides.

Acknowledgement

This material is based upon work supported by the National Science Foundation under Grant CNS-0844983. Any opinions, findings and conclusions or recomendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).