ACM Digital Library Logo
ACM Logo
Advanced Search
article

Exploiting process lifetime distributions for dynamic load balancing

Publication: ACM SIGMETRICS Performance Evaluation ReviewMay 1996 https://doi.org/10.1145/233008.233019
  • 90citation
  • 604
    • Downloads
Metrics
Total Citations90
Total Downloads604
Last 12 Months4
Last 6 weeks0

Abstract

We measure the distribution of lifetimes for UNIX processes and propose a functional form that fits this distribution well. We use this functional form to derive a policy for preemptive migration, and then use a trace-driven simulator to compare our proposed policy with other preemptive migration policies, and with a non-preemptive load balancing strategy. We find that, contrary to previous reports, the performance benefits of preemptive migration are significantly greater than those of non-preemptive migration, even when the memory-transfer cost is high. Using a model of migration costs representative of current systems, we find that preemptive migration reduces the mean delay (queueing and migration) by 35 - 50%, compared to non-preemptive migration.

References

  1. AE87.Rakesh Agrawal and Ahmed Ezzet. Location independent remote execution in NEST. IEEE Transactions on Software Engineering, 13(8):905-912, August 1987. Google ScholarGoogle Scholar
  2. AF89.Y. Artsy and R. Finkel Designing a process migration facility: The Charlotte experience. IEEE Computer, pages 47-56, September 1989. Google ScholarGoogle Scholar
  3. BF81.Raymond M. Bryant and Raphael A. Finkel. A stable distributed scheduling algorithm. In 2rid International Conference on Distributed Computing Systems, pages 314-323, 1981.Google ScholarGoogle Scholar
  4. BK90.Havio Bonomi and Anurag Kumar. Adaptive optimal load balancing in a nonhomogeneous multiserver system with a central job scheduler. IEEE Transactions on Computers, 39(10): 1232-1250, October 1990. Google ScholarGoogle Scholar
  5. Bra95.Avner Braverman, 1995. Personal Communication.Google ScholarGoogle Scholar
  6. BSW93.Amnon Barak, Guday Shai, and Richard G. Wheeler. The MOSIX Distributed Operating System:Load Balancing for UNIX Springer Vedag, Berlin, 1993. Google ScholarGoogle Scholar
  7. CCK+95.Jeremy Casas, Dan L. Clark, Ravi Konuru, Steve W. Otto, Robert M. Prouty, and Jonathan Walpole. Mpvm: A migration transparent version of pvm. Computing Systems, 8(2):171-216, Spring 1995.Google ScholarGoogle Scholar
  8. CK87.Thomas L. Casavant and Jon G. Kuhl. Analysis of three dynamic distributed load-balancing strategies with varying global informataon requirements. In 7th International Conference on Distributed Computing Systems, pages 185-192, September 1987.Google ScholarGoogle Scholar
  9. DHB95.Allen B. Downey and Mor Harchol-Balter A note on "The limited performance benefits of migrating active processes for load sharing". Technical Report UCB//CSD-95-888, University of California, Berkeley, November 1995. Google ScholarGoogle Scholar
  10. DO91.Fred Douglis and John Ousterhout. Transparent process rmgration: Design alternatives and the sprite implementation. Software- Praeuce and Experience, 21(8).757-785, August 1991. Google ScholarGoogle Scholar
  11. EB93.D.J. Evans and W U. N. Butt Dynamic load balancing using tasktransfer probablilites. Parallel Computing, 19:897-916, August 1993. Google ScholarGoogle Scholar
  12. ELZ86.Derek L. Eager, Edward D. Lazowska, and Joht~ Zahorjan. Adaptive load sharing in homogeneous distributed systems. IEEE Transactions on Software Engineering, 12(5)'662-675, May 1986. Google ScholarGoogle Scholar
  13. ELZ88.Derek L. Eager, Edward D. Lazowska, and John Zahorjan. The limited performance benefits of migrating active processes for load sharing. In S1GMETRICS, pages 662-675, May 1988. Google ScholarGoogle Scholar
  14. GGI+91.G.W. Gerrity, A. Goscinski, J. Indulska, W. Toomey, and W. Zhu. RHODOS-a testbed for studying design issues in distributed operating systems, in Towards Network Globalization (SICON 91): 2nd intert~tional Conference on Networks, pages 268-274, September 1991.Google ScholarGoogle Scholar
  15. HJ90.Anna Ha6 and Xiaowei Jin. Dynamic load balancing in a distributed system using a sender-initiated algorithm. Journal of Systems Software, 11'79-94, 1990 Google ScholarGoogle Scholar
  16. HP90.John L Hennessy and David A. Patterson. Computer Architecture A Quantitative Approach. Morgan Kaufmann Publishers, San Mateo, CA, 1990 Google ScholarGoogle Scholar
  17. KL88.Phillip Krueger and Miron Livny. A comparison of preemptive and non-preemptive load distributing In 8th International Conference on Distributed Computing Systems, pages 123-130, June 1988.Google ScholarGoogle Scholar
  18. Kun91.Thomas Kunz. The influence of different workload descriptions on a heuristic load balancing scheme. IEEE Transactions on Software Engineering, I7(7):725-730, July 1991. Google ScholarGoogle Scholar
  19. LL90.M. Litzkow and M. Livny. Experience with the Condor distributed batch system. In IEEE Workshop on Experimental Distributed Systems, pages 9%101, 1990.Google ScholarGoogle Scholar
  20. LLM88.M.J. Litzkow, M. Livny, and M.W. Mutka. Condor - a hunter of idle workstations. In 8th International Conference on Distributed Computing Systems, June 1988.Google ScholarGoogle Scholar
  21. LM82.Miron Livny and Myron Melman. Load balancing in homogeneous broadcastdistributed systems In ACM Computer Network Perfor. mance Symposium, pages 47-55, April 1982. Google ScholarGoogle Scholar
  22. LO86.W.E. Leland and T. J. Ott. Load-balancing heuristics and process behavior. In Proceedings of Performance and ACM Sigmetrics, volume 14, pages 54--69, 1986. Google ScholarGoogle Scholar
  23. LR93.Hwa-Chun Lin and C.S. Raghavendra. A state-aggregation method for analyzing dynamic load-balancing policies. In IEEE 13th International Conference on Distributed Computing Systems, pages 482--489, May 1993.Google ScholarGoogle Scholar
  24. MTS90.Raw Mirchandaney, Don Towsley, and John A. Stankovic. Adaptive load sharing in heterogeneous distributed systems. Journal of Parallel and Distributed Computing, 9:331-346, 1990. Google ScholarGoogle Scholar
  25. Nut94.Mark Nuttall. Survey of systems providing process or object migration. Technical Report DOC94/10, Imperial College Research Report, 1994.Google ScholarGoogle Scholar
  26. PM83.M.L. Powell and B.P. Miller. Process migrations in DEMOS/MP In ACM-SIGOPS 6th ACM Symposium on Operating Systems Principles, pages I 10-119, November 1983. Google ScholarGoogle Scholar
  27. PTS88.S'piridon Pulidas, Don Towsley, and John A. Stankovic. Imbedding gradient estimators in load balancing algorithms. In 8th International Conference on Distributed Computing Systems, pages 482-490, June 1988.Google ScholarGoogle Scholar
  28. Rom91.C Gary Rommel. The probability of load balancing success in a homogeneous network. IEEE Transactions on Software Engineering, 17:922-933, 1991. Google ScholarGoogle Scholar
  29. Ros65.Robert E Rosin. Determining a computing center environment. Communications of the ACM, 8(7), 1965. Google ScholarGoogle Scholar
  30. SPG94.A. Silberschatz, J.L. Peterson, and P.B. Galvin. Operating System Concepts, 4th Edition. Addison-Wesley, Reading, MA, 1994. Google ScholarGoogle Scholar
  31. Sve90.Anders Svensson. History, an intelligent load sharing filter. In IEEE lOth International Conference on Distributed Computing Systems, pages 546-553, 1990.Google ScholarGoogle Scholar
  32. Thi91.G. Thiel. Locus operating system, a transparent system. Computer Communications, 14(6):336-346, 1991. Google ScholarGoogle Scholar
  33. TLC85.Marvin M. Theimer, Keith A. Lantz, and David R Cheriton. Preemptable remote execution facilities for the V-System. In ACM-SIGOPS lOth ACM Symposium on Operating Systems Principles, pages 2-I2, December 1985. Google ScholarGoogle Scholar
  34. TvRaHvSS90.A.S. Tanenbaum, R. van Renesse adn H. van Staveren, and G.J. Sharp. Experiences with the Amoeba distributed operating system. Communications of the ACM, pages 336-346, December 1990. Google ScholarGoogle Scholar
  35. Vah95.Amin Vahdat, 1995. Personal CommunicationGoogle ScholarGoogle Scholar
  36. VGA94.Amin M. Vahdat, Douglas P. Ghormtey, and Thomas E Anderson. Eftieient,portable, and robust extension of operating system functionality. Technical Report UCB//CSD-94-842, University of California, Berkeley, 1994. Google ScholarGoogle Scholar
  37. WM85.Yung-Terng Wang and Robert J.T. Morris. Load sharing in distributed systems. IEEE Transactions on Computers, c-94(3):204-217, March 1985.Google ScholarGoogle Scholar
  38. WZKL93.J. Wang, S. Zhou, K.Ahmed, and W. Long. LSBATCH: A distributed load ~hafing batch system. Tochnical Roport C8RI-286, Computer Systems Research Institute, University of Toronto, April 1993.Google ScholarGoogle Scholar
  39. Zay87.E.R. Zayas. Attacking the process migration bottleneck. In ACM- SIGOPS llth ACM Symposium on Operating Systems Principles, pages 13-24, 1987. Google ScholarGoogle Scholar
  40. Zho89.Songnian Zhou. Performance studies for dynamic load balancing in distributed systems. PhD Dissertation, University of California, Berkeley, 1989.Google ScholarGoogle Scholar
  41. ZWZD93.S Zhou, J. Wang, X. Zheng, and P. Delisle. Utopia: a load-sharing faeitlity for large heterogeneous distributed computing systems. Software - Practice and Expeience, 23(2): 1305-1336, December 1993. Google ScholarGoogle Scholar

Index Terms

  1. Exploiting process lifetime distributions for dynamic load balancing

            Comments

            Login options

            Check if you have access through your login credentials or your institution to get full access on this article.

            Sign in

            Full Access

            Get this Article

            PDF Format

            View or Download as a PDF file.

            PDF

            eReader

            View online with eReader.

            eReader
            View Issue’s Table of Contents
            About Cookies On This Site

            We use cookies to ensure that we give you the best experience on our website.

            Learn more

            Got it!