A Practical Learning-Based Approach for Dynamic Storage Bandwidth Allocation

  • Vijay Sundaram
  • Prashant Shenoy
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2707)


In this paper, we address the problem of dynamic allocation of storage bandwidth to application classes so as to meet their response time requirements. We present an approach based on reinforcement learning to address this problem. We argue that a simple learning-based approach may not be practical since it incurs significant memory and search space overheads. To address this issue, we use application-specific knowledge to design an efficient, practical learning-based technique for dynamic storage bandwidth allocation. Our approach can react to dynamically changing workloads, provide isolation to application classes and is stable under overload. We implement our techniques into the Linux kernel and evaluate it using prototype experimentation and trace-driven simulations. Our results show that (i) the use of learning enables the storage system to reduce the number of QoS violations by a factor of 2.1 and (ii) the implementation overheads of employing such techniques in operating system kernels is small.


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    T. Abdelzaher, K.G Shin and N. Bhatti. Performance Guarantees for Web server End-Systems: AControl Theoretic Approach. IEEE Transactions on Parallel and Distributed Systems. 13(1), January 2002.Google Scholar
  2. 2.
    G. A. Alvarez et al. Minerva: An Automated Resource Provisioning Tool for Large-scale Storage Systems. ACM Transactions on Computer Systems (to appear). Technical report HPL-2001-139, Hewlett-Packard Labs, June 2001.Google Scholar
  3. 3.
    E. Anderson et al. Hippodrome: Running Circles Around Storage Administration. In FAST’02, Monterey, CA, pp. 175–188, Jan. 2002.Google Scholar
  4. 4.
    E. Anderson et al. Ergastulum: An Approach to Solving the Workload and Device Configuration Problem. HP Laboratories SSP technical memo HPL-SSP-2001-05, May 2001.Google Scholar
  5. 5.
    E. Anderson, R. Swaminathan, A. Veitch, G. A. Alvarez and J. Wilkes. Selecting RAID levels for Disk Arrays. In FAST’02, Monterey, CA, pp. 189–201, January 2002.Google Scholar
  6. 6.
    M. Aron et al. Scalable Content-aware Request Distribution in Cluster-based Network Servers. Proceedings of the USENIX 2000 Annual Technical Conference, San Diego, CA, June 2000.Google Scholar
  7. 7.
    P. Barham. A Fresh Approach to File System Quality of Service. In Proceedings of NOSSDAV’ 97, St. Louis, Missouri, pages 119–128, May 1997.Google Scholar
  8. 8.
    E. Borowsky et al. Capacity planning with phased workloads. In Proceedings of the Workshop on Software and Performance (WOSP’98), Santa Fe, NM, October 1998.Google Scholar
  9. 9.
    A. Brown, D. Oppenheimer, K. Keeton, R. Thomas, J. Kubiatowicz, and D.A. Patterson. ISTORE: Introspective Storage for Data-Intensive Network Services. In Proceedings of the 7th Workshop on Hot Topics in Operating Systems (HotOS-VII), Rio Rico, Arizona, March 1999.Google Scholar
  10. 10.
    J. Carlström and E. Nordström. Reinforcement learning for Control of Self-Similar Call Traffic in Broadband Networks. Proceedings of the 16th International Teletraffic Congress, ITC’16, P. Key., D. Smith (eds.), Elsevier Science, Edinburgh, Scotland, 1999.Google Scholar
  11. 11.
    J. Chase et al. Managing Energy and Server Resources in Hosting Centers. Proceedings of the Eighteenth ACM Symposium on Operating Systems Principles (SOSP), Oct. 2001.Google Scholar
  12. 12.
    M. Dahlin et al. A Qualitative Analysis of Cache Policies for Scalable Network File Systems. In Proceedings of the ACM SIGMETRICS’ 94, May 1994.Google Scholar
  13. 13.
    C. Lu, G. A. Alvarez, and J. Wilkes. Aqueduct: Online Data Migration with Performance Guarantees. In FAST’02, Monterey, CA, pp. 219–230, January 2002.Google Scholar
  14. 14.
    E. Nordström and J. Carlström. A Reinforcement Learning Scheme for Adaptive Link Allocation in ATM Networks. IWANNT’ 95, J. Alspector, T.X. Brown, pp. 88–95, Lawrence Erlbaum, Stockholm, Sweden, 1995.Google Scholar
  15. 15.
    D.A. Patterson et al. Recovery-Oriented Computing (ROC): Motivation, Definition, Techniques, and Case Studies. UC Berkeley Computer Science Technical Report UCB//CSD-02-1175, March 15, 2002.Google Scholar
  16. 16.
    P. Pradhan, R. Tewari, S. Sahu, A. Chandra and P. Shenoy. An Observation-based Approach Towards Self-managing Web Servers. In Proceedings of ACM/IEEE Intl Workshop on Quality of Service (IWQoS), Miami Beach, FL, May 2002.Google Scholar
  17. 17.
    D. Revel, D. McNamee, C. Pu, D. Steere and J. Walpole. Feedback Based Dynamic Proportion Allocation for Disk I/O. Technical Report CSE-99-001, OGI CSE, January 1999.Google Scholar
  18. 18.
    P. Shenoy and H. Vin. Cello: A Disk Scheduling Framework for Next Generation Operating Systems. In Proceedings of ACM SIGMETRICS’ 98, Madison, WI, pp. 44–55, June, 1998.Google Scholar
  19. 19.
    S. Singh and D. Bertsekas. Reinforcement Learning for Dynamic Channel Allocation in Cellular Telephone Systems. With D. Bertsekas. In NIPS 10, 1997.Google Scholar
  20. 20.
    V. Sundaram and P. Shenoy. Bandwidth Allocation in a Self-Managing Multimedia File Server. Proceedings of the Ninth ACM Conference on Multimedia, Ottawa, Canada, Oct. 2001.Google Scholar
  21. 21.
    R. S. Sutton and AG. Barto. Reinforcement Learning: An Introduction. MITPress, Cambridge, MA.Google Scholar
  22. 22.
    J. Ward, M. O’Sullivan, T. Shahoumian, and J. Wilkes. Appia: Automatic Storage Area Network Design. In FAST’02, Monterey, CA, pp. 203–217, January 2002.Google Scholar
  23. 23.
    R. Wijayaratne and A. L. N. Reddy. Providing QoS Guarantees for Disk I/O. Technical Report TAMU-ECE97-02, Department of Electrical Engineering, Texas A&M University, 1997.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2003

Authors and Affiliations

  • Vijay Sundaram
    • 1
  • Prashant Shenoy
    • 1
  1. 1.Department of Computer ScienceUniversity of Massachusetts AmherstUSA

Personalised recommendations