A Bayesian approach to fault classification

ABSTRACT

According to their temporal behavior, faults in computer systems are classified into permanent, intermittent, and transient faults. Since it is impossible to identify the type of a fault upon its first detection, the common practice is to retry the failed instruction one or more times and then use other fault recovery methods, such as rollback or restart, if the retry is not successful. To determine an “optimal” (in some sense) number of retries, we need to know several fault parameters, which can be estimated only after classifying all the faults detected in the past. In this paper we propose a new fault classification scheme which assigns a fault type to each detected fault based on its detection time, the outcome of retry, and its detection symptom. This classification procedure utilizes the Bayesian decision theory to sequentially update the estimation of fault parameters whenever a detected fault is classified. An important advantage of this classification is the early identification of presence of an intermittent fault so that appropriate measures can be taken before it causes a serious damage to the system. To assess the goodness of the proposed scheme, the probability of incorrect classification is also analyzed and compared with simulation results.

REFERENCES

• 1: K. G. Shin and Y.-H. Lee, "Error detection process - model, design, and its impact on computer performance," IEEE Trans. Computers, vol. C-33, pp. 529- 540, June 1984.
• 2: J. C. Laprie, "Dependable computing and fault tolerance: Concepts and terminology," Digest of papers, FTCS-15, pp. 2-11, June 1985.
• 3: P. A. Lee , T. Anderson , J. C. Laprie , A. Avizienis , H. Kopetz, Fault Tolerance: Principles and Practice, Springer-Verlag New York, Inc., Secaucus, NJ, 1990
• 4: D. P. Siewiorek and R. S. Swarz, The Theory and Practice of Reliable System Design. Bedford, MA: Digital Equipment Corporation, 1982.
• 5: D. P. Siewiorek, V. Kini, H. Mashburn, S. R. McConnel, and M. M. Tsao, "A case study of c.mmp, cm*, and c.vmp: Part i -experiences with fault tolerance in nlultiprocessor systems," Proceedings of the IEEE, vol. 66, pp. 1178-1199, Oct. 1978.
• 6: O. Tasar and V. Tasar, "A study of intermittent fault in digital computers," Proc. Nat. Comput. Con}., pp. 807- 811, June 1977.
• 7: Yann-Heng Lee , Kang G. Shin, Optimal design and use of retry in fault-tolerant computer systems, Journal of the ACM (JACM), v.35 n.1, p.45-69, Jan. 1988 [doi>10.1145/42267.42269]
• 8: M. Berg , I. Koren, On switching policies for modular redundancy fault-tolerant computing systems, IEEE Transactions on Computers, v.36 n.9, p.1052-1062, Sept. 1987 [doi>10.1109/TC.1987.5009536]
• 9: Israel Koren , Zahava Koren , Stephen Y.H. Su, Analysis of a class of recovery procedures, IEEE Transactions on Computers, v.35 n.8, p.703-712, Aug. 1986 [doi>10.1109/TC.1986.1676821]
• 10: J. O. Berger, Statistical Decision Theory, Foundations, Concepts, Methods. New York: Springer-Verlag, 2nd ed., 1985.
• 11: M. H. DeGroot, Optimal Statistical Decisions. New York: McGraw-Hill Book Company, 1970.

INDEX TERMS

Primary Classification:
  C. Computer Systems Organization
  C.4 PERFORMANCE OF SYSTEMS
      Subjects: Reliability, availability, and serviceability

Additional Classification:
  B. Hardware
  B.6 LOGIC DESIGN

  F. Theory of Computation
  F.4 MATHEMATICAL LOGIC AND FORMAL LANGUAGES
      F.4.1 Mathematical Logic
          Subjects: Computability theory

  G. Mathematics of Computing
  G.3 PROBABILITY AND STATISTICS
      Subjects: Statistical computing

General Terms:
Algorithms, Reliability, Theory