Algoritmo de Reconfiguração na Máquina T-Node em Caso de Falhas

  • Raul Ceretta Nunes UFRGS
  • Philippe Olivier Alexandre Navaux INPG / UFRGS
  • Ingrid Jansch-Pôrto INPG / UFRGS

Abstract


In this paper is presented an algorithm to the execution of reconfiguration in the T-NODE machine in presence of fault. T-NODE [TEL91] is a paralel machine which uses transputers as basic building blocks; it has been designed for high performance applications and does not present, ar its origin, any special behavior concerning fault-tolerance. In the following, we briefly present some basic concepts of fault-tolerant systems, the architecture environment of the T-NODE machine, besides explaining the motivation for the developed research. In the sequence, the main characteristics of transputers and of the T-NODE machine are described; for this description, the needs and the point-of-view of reconfiguration are considered. The algorithm to reconfigure the T-NODE under fault occurrence is then presented: detection is done considering the transputers and their local memories as basic modules. The paper finishes with the discussion of possible improvements and conclusions taken from this work.

References

DUTT, S. and HAYES, J. P. Design and Reconfiguration Strategies for Near-Optimal K-Fault-Tolerant Tree Architectures. International Symp. On Fault-Tolerant Computing, FTCS-18. New York: IEEE, 1988.

FLYNN, M. J. Very High-Speed Computing Systems. Proceedings of the IEEE, n. 54, Dezembro 1966.

HASAN, N. and LIU, L. Minimum Fault Coverage in Reconfigurable Arrays. International Symp. On Fault-Tolerant Computing, FTCS-18. New York: TEEE,1988. p.348-53

HOARE, C. A. R. Communicating Sequencial Process. Communications of the ACM. v.21, n.8, Agosto, 1978. p.666-677.

INMOS LIMITED. IMS 1424. Bristol: INMOS, 1984. 31p. (Preliminary data)

INMOS LIMITED. IMS T800 Transputer. In: Transputer databook. Bath: Bath, 1988. p.43-111

INMOS LIMITED. OCCAM 2 Reference Manual. Cambridge: Prentice Hall, 1988, 133p. (Series in Computer Science)

LAPRIE, J.C. Dependable computing and fault-tolerance: concepts and terminology. In: international Symp. On Fault-Tolerant Computing, FTCS-15. New York: IEEE, 1985. p.2-11

LOWRIE, M. and FUCHS, W. Reconfigurable Tree Architectures Using Subtree Oriented Fault Tolerance. IEEE Transactions on Computers, v. C-36, n.10, Outubro 1987, p. 1172-1182

NEGRINI, R.; SAMI, M.; STEFANELLI, R. Fault tolerance techniques for array structures used in supercomputing. Computer, v.19, n.2, Fevereiro 1986. p.78-87

NICOLE, D. A. Reconfigurable transputer processor architecture. Southampton: Southamptom Transputer Support Centre, 1988. 18p. (ESPRIT Project 1085, Tech. Report, n.2)

NUNES, R.C. Um estudo de confiabilidade da arquitetura do T-NODE. Porto Alegre: CPGCC da UFRGS, 1992. 62p. (Trabalho Individual, n.252)

POUNTAIN, D. and MAY, D. A tutorial introduction to OCCAM programming. BSP Professional Books, 1988.

SIEWIOREK, D.; SWARZ, R. The theory and practice of reliable system design. Bedford: Digital, 1982. 772p.

TELMAT INFORMATIQUE. The Configuration File for Standalone Enviroments. Technical Report n.2, Maio 1990. 9p. (In: T-NODE Technical Reports)

TELMAT INFORMATIQUE. T-NODE hardware manual. Soultz: Telmat Informatique, 1991. v.1
Published
1993-09-07
NUNES, Raul Ceretta; NAVAUX, Philippe Olivier Alexandre; JANSCH-PÔRTO, Ingrid. Algoritmo de Reconfiguração na Máquina T-Node em Caso de Falhas. In: INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE AND HIGH PERFORMANCE COMPUTING (SBAC-PAD), 5. , 1993, Florianópolis/SC. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 1993 . p. 344-357. DOI: https://doi.org/10.5753/sbac-pad.1993.23043.