Advertisement

Introduction to Multiprocessor I/O Architecture

  • David Kotz
Part of the The Kluwer International Series in Engineering and Computer Science book series (SECS, volume 362)

Abstract

The computational performance of multiprocessors continues to improve by leaps and bounds, fueled in part by rapid improvements in processor and interconnection technology. I/O performance thus becomes ever more critical, to avoid becoming the bottleneck of system performance. In this paper we provide an introduction to I/O architectural issues in multiprocessors, with a focus on disk subsystems. While we discuss examples from actual architectures and provide pointers to interesting research in the literature, we do not attempt to provide a comprehensive survey. We concentrate on a study of the architectural design issues, and the effects of different design alternatives.

Keywords

Network Interface Disk Array Disk Failure Processor Node Buffer Cache 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    George S. Almasi and Allan Gottlieb. Highly Parallel Computing. Ben-jamin/Cummings, 2nd edition, 1994.MATHGoogle Scholar
  2. [2]
    Gene M. Amdahl. Validity of the single processor approach to achieving large scale computing capabilities. In AFIPS Conference Proceedings, volume 30, pages 483–485, Spring 1967.Google Scholar
  3. [3]
    Sandra Johnson Baylor, Caroline Benveniste, and Yarsun Hsu. Performance evaluation of a massively parallel I/O subsystem. This volume.Google Scholar
  4. [4]
    Sandra Johnson Baylor, Caroline B. Benveniste, and Yarson Hsu. Performance evaluation of a parallel I/O architecture. In Proceedings of the 9th ACM International Conference on Supercomputing, pages 404–413, Barcelona, July 1995.CrossRefGoogle Scholar
  5. [5]
    BBN Advanced Computers, Cambridge, MA. Inside the Butterfly Plus, October 1987.Google Scholar
  6. [6]
    Robert E. Benner. Parallel graphics algorithms on a 1024–processor hypercube. In Proceedings of the Fourth Conference on Hypercube Concurrent Computers and Applications, pages 133–140. Golden Gate Enterprises, Los Altos, CA, 1989.Google Scholar
  7. [7]
    Lawrence Berdahl. Parallel transport protocol proposal. Lawrence Livermore National Labs, January 3, 1995. Draft.Google Scholar
  8. [8]
    Matthias A. Blumrich, Kai Li, Richard Alpert, Cezary Dubnicki, Edward W. Feiten, and Jonathan Sandberg. Virtual memory mapped network interface for the SHRIMP multicomputer. In Proceedings of the 21st Annual International Symposium on Computer Architecture, pages 142–153, 1994.Google Scholar
  9. [9]
    C. Bornstein and P. Steenkiste. Data reshuffling in support of fast I/O for distributed-memory machines. In Proceedings of the Third IEEE International Symposium on High Performance Distributed Computing, pages 227–235, August 1994.CrossRefGoogle Scholar
  10. [10]
    J. C. Browne, A. G. Dale, C. Leung, and R. Jenevein. A parallel multi-stage I/O architecture with self-managing disk cache for database management applications. In Proceedings of the Fourth International Workshop on Database Machines. Springer-Verlag, March 1985.Google Scholar
  11. [11]
    Pei Cao, Swee Boon Lim, Shivakumar Venkataraman, and John Wilkes. The TickerTAIP parallel RAID architecture. ACM Transactions on Computer Systems, 12(3): 236–269, August 1994.CrossRefGoogle Scholar
  12. [12]
    Convex Exemplar Scalable Parallel Processing System. Convex Computer Corporation, 1994. Order number 080–002293–000.Google Scholar
  13. [13]
    Peter M. Chen, Edward K. Lee, Garth A. Gibson, Randy H. Katz, and David A. Patterson. RAID: high-performance, reliable secondary storage. ACM Computing Surveys, 26(2): 145–185, June 1994.CrossRefGoogle Scholar
  14. [14]
    Tzi-cker Chiueh. Performance optimization for parallel tape arrays. In Proceedings of the 9th ACM International Conference on Supercomputing, pages 375–384, Barcelona, July 1995.CrossRefGoogle Scholar
  15. [15]
    Alok Choudhary, Rajesh Bordawekar, Michael Harry, Rakesh Krishnaiyer, Ravi Ponnusamy, Tarvinder Singh, and Rajeev Thakur. PASSION: parallel and scalable software for input-output. Technical Report SCCS-636, ECE Dept., NPAC and CASE Center, Syracuse University, September 1994.Google Scholar
  16. [16]
    Samuel S. Coleman and Richard W. Watson. New architectures to reduce I/O bottlenecks in high-performance systems. In Proceedings of the Twenty-Sixth Annual Hawaii International Conference on System Sciences, volume I, pages 5–14, 1993.CrossRefGoogle Scholar
  17. [17]
    Thomas H. Cormen and David Kotz. Integrating theory and practice in parallel file systems. In Proceedings of the 1993 DAGS/PC Symposium, pages 64–74, Hanover, NH, June 1993. Dartmouth Institute for Advanced Graduate Studies. Revised as Dartmouth PCS-TR93–188 on 9/20/94.Google Scholar
  18. [18]
    Robert A. Coyne, Harry Hulen, and Richard Watson. The high performance storage system. In Proceedings of Supercomputing ’93, pages 83–92, 1993.CrossRefGoogle Scholar
  19. [19]
    Juan Miguel del Rosario. High performance parallel I/O on the nCUBE 2. Transactions of the Institute of Electronics, Information and Communications Engineers, J75D-I(8): 626–636, August 1992.Google Scholar
  20. [20]
    Juan Miguel del Rosario and Alok Choudhary. High performance I/O for parallel computers: Problems and prospects. IEEE Computer, 27(3):59–68, March 1994.Google Scholar
  21. [21]
    David DeWitt and Jim Gray. Parallel database systems: The future of high-performance database systems. Communications of the ACM, 35(6): 85–98, June 1992.CrossRefGoogle Scholar
  22. [22]
    David J. DeWitt, Shahram Ghandeharizadeh, Donovan A. Schneider, Allan Bricker, Hui-I Hsaio, and Rick Rasmussen. The Gamma database machine project. IEEE Transactions on Knowledge and Data Engineering, 2(1):44–62, March 1990.CrossRefGoogle Scholar
  23. [23]
    Ann L. Drapeau and Randy H. Katz. Striping in large tape libraries. In Proceedings of Supercomputing ’93, pages 378–387, 1993.CrossRefGoogle Scholar
  24. [24]
    Ann L. Drapeau, Ken W. Shirrif, John H. Hartman, Ethan L. Miller, Srinivasan Seshan, Randy H. Katz, Ken Lutz, David A. Patterson, Edward K. Lee, Peter H. Chen, and Garth A. Gibson. RAID-II: a high-bandwidth network file server. In Proceedings of the 21st Annual International Symposium on Computer Architecture, pages 234–244, 1994.Google Scholar
  25. [25]
    Bob Duzett and Ron Buck. An overview of the nCUBE 3 supercomputer. In Proceedings of the Fourth Symposium on the Frontiers of Massively Parallel Computation, pages 458–464, 1992.CrossRefGoogle Scholar
  26. [26]
    Dror G. Feitelson, Peter F. Corbett, Sandra Johnson Baylor, and Yarson Hsu. Parallel I/O subsystems in massively parallel supercomputers. IEEE Parallel and Distributed Technology, pages 33–47, Fall 1995.Google Scholar
  27. [27]
    Dror G. Feitelson, Peter F. Corbett, Yarson Hsu, and Jean-Pierre Prost. Parallel I/O systems and interfaces for parallel computers. In Multiprocessor SystemsDesign and Integration. World Scientific, 1996. To appear.Google Scholar
  28. [28]
    Dror G. Feitelson, Peter F. Corbett, and Jean-Pierre Prost. Performance of the Vesta parallel file system. In Proceedings of the Ninth International Parallel Processing Symposium, pages 150–158, April 1995.CrossRefGoogle Scholar
  29. [29]
    M. J. Flynn. Very high-speed computing systems. Proceedings of the IEEE, 54(12): 1901–1909, December 1966.CrossRefGoogle Scholar
  30. [30]
    Robert J. Flynn and Haldun Hadimioglu. A distributed hypercube file system. In Proceedings of the Third Conference on Hypercube Concurrent Computers and Applications, pages 1375–1381, 1988.CrossRefGoogle Scholar
  31. [31]
    James C. French, Terrence W. Pratt, and Mriganka Das. Performance measurement of the Concurrent File System of the Intel iPSC/2 hypercube. Journal of Parallel and Distributed Computing, 17(1–2): 115–121, January and February 1993.CrossRefGoogle Scholar
  32. [32]
    Joydeep Ghosh, Kelvin D. Goveas, and Jeffrey T. Draper. Performance evaluation of a parallel I/O subsystem for hypercube multiprocessors. Journal of Parallel and Distributed Computing, 17(1–2):90–106, January and February 1993.CrossRefGoogle Scholar
  33. [33]
    Garth A. Gibson. Redundant Disk Arrays: Reliable, Parallel Secondary Storage. An ACM Distinguished Dissertation 1991. MIT Press, 1992.Google Scholar
  34. [34]
    Thomas Gross and Peter Steenkiste. Architecture implications of high-speed I/O for distributed-memory computers. In Proceedings of the 8th ACM International Conference on Supercomputing, pages 176–185, July 1994.CrossRefGoogle Scholar
  35. [35]
    Haldun Hadimioglu and Robert J. Flynn. The design and analysis of a tightly coupled hypercube file system. In Proceedings of the Fifth Annual Distributed-Memory Computer Conference, pages 1405–1410, 1990.CrossRefGoogle Scholar
  36. [36]
    R. W. Hamming. Error detecting and correcting codes. The Bell System Technical Journal, XXVI(2): 147–160, April 1950.MathSciNetGoogle Scholar
  37. [37]
    John H. Hartman and John K. Ousterhout. The Zebra striped network file system. ACM Transactions on Computer Systems, 13(3):274–310, August 1995.CrossRefGoogle Scholar
  38. [38]
    John P. Hayes, Trevor N. Mudge, Quentin F. Stout, Stephen Colley, and John Palmer. Architecture of a hypercube supercomputer. In Proceedings of the 1986 International Conference on Parallel Processing, pages 653–660, 1986.Google Scholar
  39. [39]
    W. Daniel Hillis and Lewis W. Tucker. The CM-5 connection machine: A scalable supercomputer. Communications of the ACM, 36(11):31–40, November 1993.CrossRefGoogle Scholar
  40. [40]
    David Wayne Jensen. Disk I/O In High-Performance Computing Systems. PhD thesis, Univ. Illinois, Urbana-Champagne, 1993.Google Scholar
  41. [41]
    Randy H. Katz, Garth A. Gibson, and David A. Patterson. Disk system architectures for high performance computing. Proceedings of the IEEE, 77(12): 1842–1858, December 1989.CrossRefGoogle Scholar
  42. [42]
    Michelle Y. Kim. Synchronized disk interleaving. IEEE Transactions on Computers, C-35(l l):978–988, November 1986.CrossRefGoogle Scholar
  43. [43]
    David Kotz and Ting Cai. Exploring the use of I/O nodes for computation in a MIMD multiprocessor. In IPPS ’95 Workshop on Input/Output in Parallel and Distributed Systems, pages 78–89, April 1995.Google Scholar
  44. [44]
    David Kotz and Nils Nieuwejaar. File-system workload on a scientific multiprocessor. IEEE Parallel and Distributed Technology, pages 51–60, Spring 1995.Google Scholar
  45. [45]
    Kendall Square Research technical summary. Kendall Square Research, 1992.Google Scholar
  46. [46]
    Thomas T. Kwan and Daniel A. Reed. Performance of the CM-5 scalable file system. In Proceedings of the 8th ACM International Conference on Supercom-puting, pages 156–165, July 1994.CrossRefGoogle Scholar
  47. [47]
    Charles E. Leiserson, Zahi S. Abuhamdeh, David C. Douglas, Carl R. Feyn-man, Mahesh N. Ganmukhi, Jeffrey V. Hill, Daniel Hillis, Bradley C. Kuszmaul, Margaret A. St. Pierre, David S. Wells, Monica C. Wong, Shaw-Wen Yang, and Robert Zak. The network architecture of the connection machine CM-5. In Proceedings of the Fourth Symposium on Parallel Algorithms and Architectures, pages 272–285, 1992.CrossRefGoogle Scholar
  48. [48]
    Darrell D. E. Long and Bruce R. Montague. Swift/RAID: A distributed RAID system. Computing Systems, 7(3): 333–359, Summer 1994.Google Scholar
  49. [49]
    The design of the MasPar MP-2: A cost effective massively parallel multiprocessor. MasPar Computer Corporation report number MP/P-11.92, 1992.Google Scholar
  50. [50]
    NCR 3600 product description. Technical Report ST-2119–91, NCR, San Diego, September 1991.Google Scholar
  51. [51]
    John R. Nickolls. The MasPar scalable Unix I/O system. In Proceedings of the Eighth International Parallel Processing Symposium, pages 390–394, Cancun, Mexico, April 1994.CrossRefGoogle Scholar
  52. [52]
    John R. Nickolls and Ernie Rael. Data parallel Unix input/output for a massively parallel processor. Technical Report MP/P-17.93, MasPar Computer Corporation, 1993.Google Scholar
  53. [53]
    Nils Nieuwejaar and David Kotz. Performance of the Galley parallel file system. In Fourth Workshop on Input/Output in Parallel and Distributed Systems, May 1996.Google Scholar
  54. [54]
    David Patterson, Garth Gibson, and Randy Katz. A case for redundant arrays of inexpensive disks (RAID). In Proceedings of the ACM SIGMOD International Conference on Management of Data, pages 109–116, June 1988.Google Scholar
  55. [55]
    David A. Patterson and John L. Hennessy. Computer Architecture: A Quantitative Approach. Morgan Kaufmann, 1990.Google Scholar
  56. [56]
    Paul Pierce. A concurrent file system for a highly parallel mass storage system. In Proceedings of the Fourth Conference on Hypercube Concurrent Computers and Applications, pages 155–160. Golden Gate Enterprises, Los Altos, CA, March 1989.Google Scholar
  57. [57]
    Terrence W. Pratt, James C. French, Phillip M. Dickens, and Stanley A. Janet, Jr. A comparison of the architecture and performance of two parallel file systems. In Proceedings of the Fourth Conference on Hypercube Concurrent Computers and Applications, pages 161–166. Golden Gate Enterprises, Los Altos, CA, 1989.Google Scholar
  58. [58]
    A. L. Reddy, P. Banerjee, and Santosh G. Abraham. I/O embedding in hypercubes. In Proceedings of the 1988 International Conference on Parallel Processing, volume 1, pages 331–338, 1988.Google Scholar
  59. [59]
    Andrew P. Russo. The AlphaServer 2100 I/O subsystem. Digital Technical Journal, 6(3):20–28, Summer 1994.Google Scholar
  60. [60]
    Kenneth Salem and Hector Garcia-Molina. Disk striping. In Proceedings of the IEEE 1986 Conference on Data Engineering, pages 336–342, 1986.Google Scholar
  61. [61]
    Mark Smotherman. A sequencing-based taxonomy of I/O systems and review of historical machines. Computer Architecture News, 17(5):5–15, September 1989.CrossRefGoogle Scholar
  62. [62]
    Michael Stonebraker and Gerhard A. Schloss. Distributed RAID — A new multiple copy algorithm. In Proceedings of 6th International Data Engineering Conference, pages 430–437, 1990.CrossRefGoogle Scholar
  63. [63]
    DBC/1012. Teradata Corporation Booklet, 1988.Google Scholar
  64. [64]
    Thinking Machines Corporation. Programming the CM I/O System, November 1990.Google Scholar
  65. [65]
    The Connection Machine CM-5 Technical Summary, Thinking Machines Corporation, October 1991.Google Scholar
  66. [66]
    The CM-5 I/O system. Thinking Machines Corporation glossy, 1993.Google Scholar
  67. [67]
    Thorsten von Eicken, Anindya Basu, Vineet Buch, and Werner Vogels. U-Net: A user-level network interface for parallel and distributed computing. In Proceedings of the Fifteenth ACM Symposium on Operating Systems Principles, pages 40–53, December 1995.CrossRefGoogle Scholar
  68. [68]
    John Wilkes. DataMesh, house-building, and distributed systems technology. ACM Operating Systems Review, 27(2): 104–108, April 1993.CrossRefGoogle Scholar
  69. [69]
    John Wilkes, Richard Golding, Carl Staelin, and Tim Sullivan. The HP AutoRAID hierarchical storage system. In Proceedings of the Fifteenth ACM Symposium on Operating Systems Principles, pages 96–108, December 1995.CrossRefGoogle Scholar
  70. [70]
    Andrew Witkowski, Kumar Chandrakumar, and Greg Macchio. Concurrent I/O system for the hypercube multiprocessor. In Proceedings of the Third Conference on Hypercube Concurrent Computers and Applications, pages 1398–1407, 1988.CrossRefGoogle Scholar

Copyright information

© Kluwer Academic Publishers 1996

Authors and Affiliations

  • David Kotz
    • 1
  1. 1.Department of Computer ScienceDartmouth CollegeHanoverUSA

Personalised recommendations