Um Sistema de Arquivos Distribuídos Tolerante a Falhas para UNIX Compatível com o NFS
Resumo
É apresentado o projeto e a implementação de um sistema de arquivos distribuídos para o UNIX, voltado à obtenção de alta confiabilidade no armazenamento de dados, além de alta disponibilidade de acesso. Foi usado como base o sistema de arquivos NFS (Network File System), estendido para tolerar falhas através da replicação de arquivos e diretórios entre diversos servidores. O método de replicação empregado é baseado em um algoritmo de cópia principal, com distribuição síncrona de atualizações do servidor principal aos servidores secundários. Os clientes do sistema são capazes de chavear automaticamente o servidor no caso de falhas, com continuidade no acesso aos dados. Requisições de acesso por parte dos clientes levam à eleição de novo servidor principal, em caso colapso do principal, com proteções contra falhas de particionamento de rede. É proposta a recuperação de volumes em servidores religados de forma automática.
Referências
BHIDE, A.; ELNOZAHY, E.; MORGAN, S. lmplicit Replication in a Network File Server. In: WORKSHOP ON MANAGEMENT OF REPLICATED DATA, 1990, Houston. Washington: IEEE Computer Society Press, c1990.
BRERETON. O. P. Management of Replicated Files in a UNIX Environment. Software - Practice and Experience, Sussex, England, v. 16, p.77 1-780, Aug. 1986.
DAVIDSON, S. B.; GARCIA-MOLINA, H. Consistency in Partitioned Networks. ACM Computing Surveys, New York, v. 17, n. 3, p.341-370, Sept. 1985.
EL-ABBADI, A; SKEEN, A. D; CRISTIAN, F. An Efficient Fault-Tolerant Protocol for Replicated Data Management. In: SYMP. ON PRINCIPLES OF DATABASE SYSTEMS, 4., [S.I.], 1985. Proceedings ... NY ACM Press, 1985. p.215-229.
GARCIA-MOLINA, H. Elections in a Distributed Computing System. IEEE Trans. on Computers, New York, v. C-31, n. 1, p.48-59, Jan. 1982.
GIFFORD, D. K. Weighted Voting for Replicated Data. In: SYMPOSIUM ON OPERATING SYSTEMS PRINCIPLES, 7., 1979, New York. Proceedings ... New York; ACM Press, 1979. p.I50-162.
HISGEN, A. et al. Granularity and Semantic Level of Replication in the Echo Distributed File System. In: WORKSHOP ON MANAGEMENT OF REPLICATED DATA, 1990, Houston. Washington: IEEE Computer Society Press, c1990.
HUANG, Y.; JALOTE, P. Avaliability Analysis of the Primary Site Approach for Fault Tolerance. Acta Informatica, Berlin, n. 26, p.543-557, 1989.
JALOTE, Pankaj. Fault Tolerance in Distributed Systems. Englewood Cliffs, NJ.: Prentice-Hall, 1994. 432p.
KLEIMAN, S. Vnodes: An Architecture for Multiple File System Types in SUN UNIX. In: USENIX SUMMER '86 CONFERENCE, 1986. [S.I.: s.n.], p. 238-247.
KUMAR, A. Hierarchical Quorum Consensus: A New Alghoritm for Managing Replicated Data. IEEE Trans. on Computers, New York, v. 40, n. 9, p.996-1004, Sept. 1994.
LADIN, Rivka; LISKOV, Barbara; SHRIRA, Liuba. Lazy Replication: Exploiting the Semantics of Distributcd Services. Operating Systems Review, New York, v.25, n.1, p.49-55, Jan. 1991.
LEVY, E; SILBERSCHATZ, A. Distributed File Systems: Concepts and Examples. Computing Surveys, New York, v. 22, n. 4, p.321-374, Dec. 1990.
LISKOV, Barbara et al. Replication in the Harp File System. Operating Systems Review, New York, 1991, v. 25, n 5, p. 226-238.
LISKOV, Barbara et al. A Replicated Unix File System. Operating Systems Review, New York, v.25, n. 1, p.60-64, Jan. 1991.
NELSON, M. N; WELCH, B. B; OUSTERHOUT, J.K. Caching in the Sprite Network File System. ACM Trans. on Computer Systems, New York, v. 6, n.1, p. 134-154, Feb. 1988.
PURDIN, T. O. M.; SCHLICHTING, R. D.; ANDREWS, G.R. A File Replication Facility for Berkeley UNIX. Software - Practice and Experience, Sussex, England, v.17, p.923-940, Dec. 1987.
IETF Requesl For Comments 1014. XDR: External Data Representation Standard. Disponível no servidor https://www.internic.net.
IETF Request For Comments 1057. SUN RPC Protocol Especification. Disponível no servidor https://www.internic.net.
IETF Request For Comments 1094. The NFS Protocol Especification. Disponível no servidor https://www.internic.net.
IETF Requesl For Comments 1813. NFS Version 3 Protocol. Disponível no servidor https://www.internic.net.
SANDBERG, R. et al. Design and lmplementation of the Sun Network File System. In: USENIX ASSOCIATION CONFERENCE, 1985, Berkeley. Proceedings ... Berkeley: USENIX, 1985.
SOTOMA, IRINEU; WEBER, TAISY SILVA. AFIDS- Arquitetura para Injeção de Falhas em Sistemas Distribuídos. In: Anais do 15 SIMPÓSIO BRASILEIRO DE REDES DE COMPUTADORES- XV SBRC. UFScar- Universidade Federal de São Carlos. 19-22 de maio de 1997. São Carlos - SP. p294-309.
SRINIVASAN, V; MOGUL, J. C. Spiritely NFS: Experiments with Cacheconsistency Protocols. Operating Systems Review, New York, v. 23, n.5, p. 45-57, Dec. 1989.
STERN, Hal. Managing NFS and NIS. Sebastopol, Ca: O'Reilly & Associates. 1991. 410p.
TANENBAUM, Andrew S. Modern Operating Systems. Englewood Cliffs, NJ. Prentice-Hall, 1992. 728p.
BRERETON. O. P. Management of Replicated Files in a UNIX Environment. Software - Practice and Experience, Sussex, England, v. 16, p.77 1-780, Aug. 1986.
DAVIDSON, S. B.; GARCIA-MOLINA, H. Consistency in Partitioned Networks. ACM Computing Surveys, New York, v. 17, n. 3, p.341-370, Sept. 1985.
EL-ABBADI, A; SKEEN, A. D; CRISTIAN, F. An Efficient Fault-Tolerant Protocol for Replicated Data Management. In: SYMP. ON PRINCIPLES OF DATABASE SYSTEMS, 4., [S.I.], 1985. Proceedings ... NY ACM Press, 1985. p.215-229.
GARCIA-MOLINA, H. Elections in a Distributed Computing System. IEEE Trans. on Computers, New York, v. C-31, n. 1, p.48-59, Jan. 1982.
GIFFORD, D. K. Weighted Voting for Replicated Data. In: SYMPOSIUM ON OPERATING SYSTEMS PRINCIPLES, 7., 1979, New York. Proceedings ... New York; ACM Press, 1979. p.I50-162.
HISGEN, A. et al. Granularity and Semantic Level of Replication in the Echo Distributed File System. In: WORKSHOP ON MANAGEMENT OF REPLICATED DATA, 1990, Houston. Washington: IEEE Computer Society Press, c1990.
HUANG, Y.; JALOTE, P. Avaliability Analysis of the Primary Site Approach for Fault Tolerance. Acta Informatica, Berlin, n. 26, p.543-557, 1989.
JALOTE, Pankaj. Fault Tolerance in Distributed Systems. Englewood Cliffs, NJ.: Prentice-Hall, 1994. 432p.
KLEIMAN, S. Vnodes: An Architecture for Multiple File System Types in SUN UNIX. In: USENIX SUMMER '86 CONFERENCE, 1986. [S.I.: s.n.], p. 238-247.
KUMAR, A. Hierarchical Quorum Consensus: A New Alghoritm for Managing Replicated Data. IEEE Trans. on Computers, New York, v. 40, n. 9, p.996-1004, Sept. 1994.
LADIN, Rivka; LISKOV, Barbara; SHRIRA, Liuba. Lazy Replication: Exploiting the Semantics of Distributcd Services. Operating Systems Review, New York, v.25, n.1, p.49-55, Jan. 1991.
LEVY, E; SILBERSCHATZ, A. Distributed File Systems: Concepts and Examples. Computing Surveys, New York, v. 22, n. 4, p.321-374, Dec. 1990.
LISKOV, Barbara et al. Replication in the Harp File System. Operating Systems Review, New York, 1991, v. 25, n 5, p. 226-238.
LISKOV, Barbara et al. A Replicated Unix File System. Operating Systems Review, New York, v.25, n. 1, p.60-64, Jan. 1991.
NELSON, M. N; WELCH, B. B; OUSTERHOUT, J.K. Caching in the Sprite Network File System. ACM Trans. on Computer Systems, New York, v. 6, n.1, p. 134-154, Feb. 1988.
PURDIN, T. O. M.; SCHLICHTING, R. D.; ANDREWS, G.R. A File Replication Facility for Berkeley UNIX. Software - Practice and Experience, Sussex, England, v.17, p.923-940, Dec. 1987.
IETF Requesl For Comments 1014. XDR: External Data Representation Standard. Disponível no servidor https://www.internic.net.
IETF Request For Comments 1057. SUN RPC Protocol Especification. Disponível no servidor https://www.internic.net.
IETF Request For Comments 1094. The NFS Protocol Especification. Disponível no servidor https://www.internic.net.
IETF Requesl For Comments 1813. NFS Version 3 Protocol. Disponível no servidor https://www.internic.net.
SANDBERG, R. et al. Design and lmplementation of the Sun Network File System. In: USENIX ASSOCIATION CONFERENCE, 1985, Berkeley. Proceedings ... Berkeley: USENIX, 1985.
SOTOMA, IRINEU; WEBER, TAISY SILVA. AFIDS- Arquitetura para Injeção de Falhas em Sistemas Distribuídos. In: Anais do 15 SIMPÓSIO BRASILEIRO DE REDES DE COMPUTADORES- XV SBRC. UFScar- Universidade Federal de São Carlos. 19-22 de maio de 1997. São Carlos - SP. p294-309.
SRINIVASAN, V; MOGUL, J. C. Spiritely NFS: Experiments with Cacheconsistency Protocols. Operating Systems Review, New York, v. 23, n.5, p. 45-57, Dec. 1989.
STERN, Hal. Managing NFS and NIS. Sebastopol, Ca: O'Reilly & Associates. 1991. 410p.
TANENBAUM, Andrew S. Modern Operating Systems. Englewood Cliffs, NJ. Prentice-Hall, 1992. 728p.
Publicado
07/10/1997
Como Citar
LEBOUTE, Mário Magalhães; WEBER, Taisy Silva.
Um Sistema de Arquivos Distribuídos Tolerante a Falhas para UNIX Compatível com o NFS. In: INTERNATIONAL SYMPOSIUM ON COMPUTER ARCHITECTURE AND HIGH PERFORMANCE COMPUTING (SBAC-PAD), 9. , 1997, Campos do Jordão/SP.
Anais [...].
Porto Alegre: Sociedade Brasileira de Computação,
1997
.
p. 167-183.
DOI: https://doi.org/10.5753/sbac-pad.1997.22623.
