Dissecting distributed coordination
This paper derives necessary and sufficient communication for distributed applications that perform certain actions uniformly in asynchronous systems. We show there is an essential structure of information flow in any solution to Uniform Coordination, suggesting message-minimal solutions. We show it is necessary for processes to conspire against each other to make progress, and we show this conspiracy requires processes to stop communicating with each other. This, we show, renders Uniform Coordination insensitive to channel delivery guarantees. We introduce the notion of exempting processes from coordinating. We show that ‘primary partition’ behavior (Isis) arises from the desire to make exempt an process indistinguishable from a crashed process. Defining weaker exemptions for distributed coordination problems gives rise to many problems solvable in asynchronous systems as well as in systems that partition.
KeywordsAgreement asynchronous exemption failures uniformity
Unable to display preview. Download preview PDF.
- 1.Y. Amir, D. Dolev, S. Kramer, and D. Malki. Transis: A Communication Sub-System for High Availability. In 22nd FTCS, pages 76–84. IEEE, 1992.Google Scholar
- 2.K. P. Birman and R. van Renesse. Reliable Distributed Computing with the Isis Toolkit. IEEE Computer Society Press, 1994.Google Scholar
- 3.T. D. Chandra, V. Hadzilacos, and S. Toueg. The Weakest Failure Detector for Solving Consensus. In 11th PODC, pages 147–158. ACM, 1992.Google Scholar
- 6.B. Chor and C. Dwork. Randomization in Byzantine Agreement. Advances in Computer Research, 5:443–497, 1989.Google Scholar
- 8.A. Gopal and S. Toueg. Reliable Broadcast in Synchronous and Asynchronous Environments. In 3rd WDAG-LNCS 392, pages 110–123. Springer Verlag, 1989.Google Scholar
- 12.D. Malki, K. P. Birman, A. Ricciardi, and A. Schiper. Uniform Actions in Asynchronous Distributed Systems. In 13th PODC, pages 274–284. ACM, 1994.Google Scholar
- 13.A. Ricciardi. Dissecting Distributed Coordination in Asynchronous Systems. Technical Report ECE-PDS-9502, The Univerisity of Texas, 1995.Google Scholar
- 14.A. Ricciardi. Sequential Distributed Coordination. Technical Report ECE-PDS-9503, The Univerisity of Texas, 1995.Google Scholar
- 15.A. Ricciardi and K. Birman. Using Process Groups to Implement Failure Detection in Asynchronous Environments. In 10th PODC, pages 341–351. ACM, 1991.Google Scholar
- 16.L. Sabel and K. Marzullo. Simulating Fail-Stop in Asynchronous Distributed Systems. In 13th Symposium on Rel. Dist. Sys., pages 138–47. IEEE, 1994.Google Scholar
- 17.R. van Renesse, T. Hickey, and K. Birman. Design and Performance of Horus: A Lightweight Group Communications System. Technical Report 94-1441, Cornell University, 1994.Google Scholar