Algorithmica

, Volume 56, Issue 4, pp 550–576 | Cite as

A Timing Assumption and Two t-Resilient Protocols for Implementing an Eventual Leader Service in Asynchronous Shared Memory Systems

  • Antonio Fernández
  • Ernesto Jiménez
  • Michel Raynal
  • Gilles Trédan
Article

Abstract

This paper considers the problem of electing an eventual leader in an asynchronous shared memory system. While this problem has received a lot of attention in message-passing systems, very few solutions have been proposed for shared memory systems. As an eventual leader cannot be elected in a pure asynchronous system prone to process crashes, the paper first proposes to enrich the asynchronous system model with an additional assumption. That assumption (denoted AWB) is particularly weak. It is made up of two complementary parts. More precisely, it requires that, after some time, (1) there is a process whose write accesses to some shared variables be timely, and (2) the timers of (tf) other processes be asymptotically well-behaved (t denotes the maximal number of processes that may crash, and f the actual number of process crashes in a run). The asymptotically well-behaved timer notion is a new notion that generalizes and weakens the traditional notion of timers whose durations are required to monotonically increase when the values they are set to increase (a timer works incorrectly when it expires at arbitrary times, i.e., independently of the value it has been set to).

The paper then focuses on the design of t-resilient AWB-based eventual leader protocols. “t-resilient” means that each protocol can cope with up to t process crashes (taking t=n−1 provides wait-free protocols, i.e., protocols that can cope with any number of process failures). Two protocols are presented. The first enjoys the following noteworthy properties: after some time only the elected leader has to write the shared memory, and all but one shared variables have a bounded domain, be the execution finite or infinite. This protocol is consequently optimal with respect to the number of processes that have to write the shared memory. The second protocol guarantees that all the shared variables have a bounded domain. This is obtained at the following additional price: t+1 processes are required to forever write the shared memory. A theorem is proved which states that this price has to be paid by any protocol that elects an eventual leader in a bounded shared memory model. This second protocol is consequently optimal with respect to the number of processes that have to write in such a constrained memory model. In a very interesting way, these protocols show an inherent tradeoff relating the number of processes that have to write the shared memory and the bounded/unbounded attribute of that memory.

Keywords

Asynchronous system Atomic register Eventual leader Fault-tolerance Omega Process crash Shared memory System model Timer property Timing assumptions t-resilient protocol 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Abraham, I., Chockler, G.V., Keidar, I., Malkhi, D.: Byzantine disk Paxos, optimal resilience with Byzantine shared memory. In: Proc. 23th ACM Symposium on Principles of Distributed Computing (PODC’04), pp. 226–235. ACM Press, New York (2004) Google Scholar
  2. 2.
    Aguilera, M.K., Delporte-Gallet, C., Fauconnier, H., Toueg, S.: On implementing omega with weak reliability and synchrony assumptions. In: Proc. 22th ACM Symposium on Principles of Distributed Computing (PODC’03), pp. 306–314. ACM Press, New York (2003) Google Scholar
  3. 3.
    Aguilera, M.K., Delporte-Gallet, C., Fauconnier, H., Toueg, S.: Communication-efficient leader election and consensus with limited link synchrony. In: Proc. 23th ACM Symposium on Principles of Distributed Computing (PODC’04), pp. 328–337. ACM Press, New York (2004) Google Scholar
  4. 4.
    Aguilera, M.K., Englert, B., Gafni, E.: On using network attached disks as shared memory. In: Proc. 21th ACM Symposium on Principles of Distributed Computing (PODC’03), pp. 315–324. ACM Press, New York (2003) Google Scholar
  5. 5.
    Chandra, T.D., Toueg, S.: Unreliable failure detectors for reliable distributed systems. J. ACM 43(2), 225–267 (1996) MATHCrossRefMathSciNetGoogle Scholar
  6. 6.
    Chandra, T.D., Hadzilacos, V., Toueg, S.: The weakest failure detector for solving consensus. J. ACM 43(4), 685–722 (1996) MATHCrossRefMathSciNetGoogle Scholar
  7. 7.
    Dwork, C., Lynch, N., Stockmeyer, L.: Consensus in the presence of partial synchrony. J. ACM 35(2), 288–323 (1988) CrossRefMathSciNetGoogle Scholar
  8. 8.
    Fernández, A., Jiménez, E., Raynal, M.: Electing an eventual leader in an asynchronous shared memory system. In: Proc. 37th International IEEE Conference on Dependable Systems and Networks (DSN’07), pp. 399–408. IEEE Computer Society Press, Los Alamitos (2007) CrossRefGoogle Scholar
  9. 9.
    Gafni, E., Lamport, L.: Disk Paxos. Distrib. Comput. 16(1), 1–20 (2003) CrossRefGoogle Scholar
  10. 10.
    Gibson, G.A., Nagle, D., Amiri, K., Butler, J., Chang, F.W., Gobioff, H., Hardin, C., Riedel, E., Rochberg, D., Zelenka, J.: A cost-effective high-bandwidth storage architecture. In: Proc. 8th International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS’98), pp. 92–103. ACM Press, New York (1998) CrossRefGoogle Scholar
  11. 11.
    Guerraoui, R., Kapalka, M., Kouznetsov, P.: The weakest failure detectors to boost obstruction-freedom. In: Proc. 20th Symposium on Distributed Computing (DISC’06). Lecture Notes in Computer Science, vol. 4167, pp. 376–390. Springer, Berlin (2006) Google Scholar
  12. 12.
    Guerraoui, R., Raynal, M.: The information structure of indulgent consensus. IEEE Trans. Comput. 53(4), 453–466 (2004) CrossRefGoogle Scholar
  13. 13.
    Guerraoui, R., Raynal, M.: A leader election protocol for eventually synchronous shared memory systems. In: 4th International IEEE Workshop on Software Technologies for Future Embedded and Ubiquitous Systems (SEUS’06), pp. 75–80. IEEE Computer Society Press, Los Alamitos (2006) CrossRefGoogle Scholar
  14. 14.
    Guerraoui, R., Raynal, M.: The alpha of asynchronous consensus. Comput. J. 50(1), 53–67 (2007) CrossRefGoogle Scholar
  15. 15.
    Herlihy, M.P.: Wait-free synchronization. ACM Trans. Program. Lang. Syst. 11(1), 124–149 (1991) CrossRefGoogle Scholar
  16. 16.
    Herlihy, M.P., Luchangco, V., Moir, M.: Obstruction-free synchronization: double-ended queues as an example. In: Proc. 23th IEEE International Conference on Distributed Computing Systems (ICDCS’03), pp. 522–529. IEEE Computer Society Press, Los Alamitos (2003) Google Scholar
  17. 17.
    Herlihy, M.P., Luchangco, V., Moir, M., Scherer III, W.N.: Software transactional memory for dynamic sized data structure. In: Proc. 21th ACM Symposium on Principles of Distributed Computing (PODC’03), pp. 92–101. ACM Press, New York (2003) Google Scholar
  18. 18.
    Herlihy, M.P., Wing, J.M.: Linearizability: a correctness condition for concurrent objects. ACM Trans. Program. Lang. Syst. 12(3), 463–492 (1990) CrossRefGoogle Scholar
  19. 19.
    Lamport, L.: The part-time parliament. ACM Trans. Comput. Syst. 16(2), 133–169 (1998) (the first version of Paxos appeared a DEC Tech Report in 1989) CrossRefGoogle Scholar
  20. 20.
    Larrea, M., Fernández, A., Arévalo, S.: Optimal implementation of the weakest failure detector for solving consensus. In: Proc. 19th Symposium on Resilient Distributed Systems (SRDS’00), pp. 52–60. IEEE Computer Society Press, Los Alamitos (2000) CrossRefGoogle Scholar
  21. 21.
    Lee, E.K., Thekkath, C.: Petal: distributed virtual disks. In: Proc. 7th International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS’96), pp. 84–92. ACM Press, New York (1996) CrossRefGoogle Scholar
  22. 22.
    Lo, W.-K., Hadzilacos, V.: Using failure detectors to solve consensus in asynchronous shared memory systems. In: Proc. 8th International Workshop on Distributed Computing (WDAG’94). Lecture Notes in Computer Science, vol. 857, pp. 280–295. Springer, Berlin (1994) Google Scholar
  23. 23.
    Malkhi, D., Oprea, F., Zhou, L.: Ω meets Paxos: leader election and stability without eventual timely links. In: Proc. 19th International Symposium on Distributed Computing (DISC’05). Lecture Notes in Computer Science, vol. 3724, pp. 199–213. Springer, Berlin (2005) Google Scholar
  24. 24.
    Mills, D.L.: Network Time Protocol (Version 3). Request for Comments (RFC) 1305, March 1992 Google Scholar
  25. 25.
    Mostefaoui, A., Mourgaya, E., Raynal, M.: Asynchronous implementation of failure detectors. In: Proc. International IEEE Conference on Dependable Systems and Networks (DSN’03), pp. 351–360. IEEE Society Press, Los Alamitos (2003) CrossRefGoogle Scholar
  26. 26.
    Mostefaoui, A., Raynal, M.: Leader-based consensus. Parallel Process. Lett. 11(1), 95–107 (2001) CrossRefMathSciNetGoogle Scholar
  27. 27.
    Mostéfaoui, A., Mourgaya, E., Raynal, M., Travers, C.: Time-free assumption to implement eventual leadership. Parallel Process. Lett. 16(2), 189–208 (2006) CrossRefMathSciNetGoogle Scholar
  28. 28.
    Mostéfaoui, A., Raynal, M., Travers, C.: Time-free and timeliness assumptions can be combined to get eventual leadership. IEEE Trans. Parallel Distrib. Syst. 17(7), 656–666 (2006) CrossRefGoogle Scholar
  29. 29.
    Powell, D.: Failure mode assumptions and assumption coverage. In: Proc. of the 22nd International Symposium on Fault-Tolerant Computing (FTCS-22), pp. 386–395. IEEE Computer Society Press, Boston (1992) Google Scholar
  30. 30.
    Raynal, M.: A short introduction to failure detectors for asynchronous distributed systems. ACM SIGACT News, Distrib. Comput. Column 36(1), 53–70 (2005) CrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2008

Authors and Affiliations

  • Antonio Fernández
    • 1
  • Ernesto Jiménez
    • 2
  • Michel Raynal
    • 3
  • Gilles Trédan
    • 3
  1. 1.LADyR, GSyCUniversidad Rey Juan CarlosMóstolesSpain
  2. 2.EUIUniversidad Politécnica de MadridMadridSpain
  3. 3.IRISAUniversité de RennesRennesFrance

Personalised recommendations