Leases: an efficient fault-tolerant mechanism for distributed file cache consistency

ABSTRACT

Caching introduces the overhead and complexity of ensuring consistency, reducing some of its performance benefits. In a distributed system, caching must deal with the additional complications of communication and host failures.

Leases are proposed as a time-based mechanism that provides efficient consistent access to cached data in distributed systems. Non-Byzantine failures affect performance, not correctness, with their effect minimized by short leases. An analytic model and an evaluation for file access in the V system show that leases of short duration provide good performance. The impact of leases on performance grows more significant in systems of larger scale and higher processor performance.

References

1. 1.BACH, M. J., LUPpI, M. W., MELAMED, A. 8., AND YUEH, K. A remote-file cache for RFS. In Proceedings of the Summer 1987 Usenix Conference (June 1987), Usenix Association, pp. 273-279.]]
2. 2.BURROWS, M. Efficient data sharing. Tech. Rep. No. 153, Computer Laboratory, University of Cambridge, Dec. I988. The author's PhD thesis.]]
3. 3.CHERrrON, D., G(X)SEN, H., AND BOYLE, P. Multilevel shared caching techniques for scalability in VMP-MC. In Proc. f 6th Int. Syrup. on Computer Architecture (May 1989).]]
4. 4.Ct3ERiTON, D. R. The V distributed system. Commun. ACM 31, 3 (Mar. 1988), 314-333.]]
5. 5.CHERITON, D. R., AND DEERING, S. E. Host groups: A multicast extension for datagram intemetworks. In Proc. 9th Data Communications Symposium (Sept. 1985), ACM/IEEE, pp. 172--179.]]
6. 6.CimmTON, D. R., AND ZWAENEPOEL, W. Distributed process groups in the V kernel. ACM Trans. Comput. Syst. 3, 2 (May 1985), 77-107.]]
7. 7.FLEIsCH, B. D., AND POPEK, G. J. Mirage: A coherent distributed shared memory design. In Proceedings of the Twelfth ACM Symposium on Operating Systems Principles (Dec. 1989), ACM.]]
8. 8.FLOYD, R. Short-term file reference patterns in a UNIX environment. Tech. Rep. TR 177, University of Rochester, Department of Computer Science, Mar. 1986.]]
9. 9.GRAY, C. G. Performance and Fault-Tolerance in a Cache for Distributed File Service. PhD thesis; Stanford University, Department of Computer Science, 1989. in preparation.]]
10. 10.HOWARD, J. H., KAZAR, M. L., ME~s, S. G., NtCHOLS, D. A., $ATYANARAYANAN, 1%/I., SIDE- BOTHAM, R. N., AND WEST, M. I. Scale and performance in a distributed file system. ACM Trans. Comput. Syst. 6, 1 (Feb. 1988), 51-81.]]
11. 11.KAZAR, M. L. Synchronization and caching issues in the Andrew file system. Tech. Rep. CMU-iTC- 058, Information Technology Center, Carm~gie Mellon Umversity, June 1987.]]
12. 12.LAMPSON, B. W. Designing a global name service. In Proceedings of the Fifth AnnuaI ACM Symposium on the Principles of Distributed Computing (Aug. 1986), ACM, pp. 1-10.]]
13. 13.MANN, T, H;SOEN, A., AND SWART, G. An algorithm for data replication. Research Report 46, DEC Systems Research Center, 1989.]]
14. 14.MOCKAPETRIS, P. Domain names --- concx~pts and facilities. Request for Comments 1034, Network information Center, SRI International, Menlo Park, CA, Nov. 1987.]]
15. 15.MONTOOMERY, W. Measurements of sharing in MULTICS. In Proc. of Sixth ACM Symposium on Operating Systems Principles (1977), ACM, pp. 85- 90.]]
16. 16.NELSON, M. N., WELCH, B. B., AND OUSTERHoLrr, J. K. Caching in the Sprite network file system. ACM Trans. Comput. Syst. 6, 1 (Feb. 1988), 134--154.]]
17. 17.OUSTERI-IOLrr, J. K., COSTA, H. D., HARRISON, D., KUNZE, J. A., KLrPP'ER, M., AND THOMPSON, J. G. A trace-driven analysis of the UNIX 4.2BSD file system. In Proceedings of the Tenth ACM Symposium on Operating Systems Principles (Dec. 1985), ACM, pp. 15-24. Published as Operating Systems Review 19,5.]]
18. 18.SATYANAR3%YANAN, M., HOWARD, J. H., NXCHOLS, D. A., SIDEBOTHAM, R. N., SPECTOR, A. Z., AND WEST, M. J. The ITC distributed file system: Principles and design. In Proceedings of the Tenth ACM Symposium on Operating Systems Principles (Dec. 1985), ACM, pp. 35-50. Published as Operating Systems Review 19, 5.]]
19. 19.SCVmOEDER, M. D., G~rORD, D. K., AND NEED- ~AM, R.M. A caching file system for a programmer's workstation, in Proceedings of the Tenth ACM Symposium on Operating Systems Principles (Dec. 1985), ACM, pp. 25-34. Published as Operating Systems Review 19, 5.]]
20. 20.STUOIS, H., MrrCHELL, J., AND ISRAEL, J. Issues in the design and use of a distributed file system. (pp. crating Systems Rm,iew 24, 3 (July 1980), 55-69.]]
21. 21.SUN MiCROSYSTEMS, INC. SunOS Reference Manual, 1988.]]
22. 22.TERRY, D.B. Distributed name servers: Naming and caching in large distributed computing environments. Tech. Rep. UCB/CSD 85/228, Computer Science Division (EECS), University of California, Mar. 1985. The author's Phl) thesis.]]
23. 23.TERRY, D. B. Caching hints in distributed systems. IEEE Trans. SoftTM. Eng. SE-13, I (Jan. 1987), 48- 54.]]
24. 24.T~OMPSON, J. G. Efficient analysis of caching systems. Tech. Rep. UCB/CSD 87/374, Computer Science Division (EECS), University of California, Oct. 1987. The author's PhD thesis.]]

Index Terms

• General and reference

  • Cross-computing tools and techniques

    • Performance

• Software and its engineering

  • Software organization and properties

    • Contextual software domains

      • Operating systems

        • Memory management

          • Distributed memory

    • Extra-functional properties

      • Software fault tolerance

      • Software performance

    • Software system structures

      • Distributed systems organizing principles