Robust asynchronous algorithms in networks with a fault detection ring
In this paper we study asynchronous networks, in which processors may fail and recover several times during an execution. Upon recovery a processor rejoins the execution as a passive processor, with all variables initialized, except for a fault flag, which is updated by the system. In addition to the regular communication links connecting pairs of processors, the processors are connected by a unidirectional ring of communication links, such that upon a failure of a processor all of the processor's incoming messages are transferred to the outgoing link belonging to this ring. We present a general transformation, that modifies any algorithm, that is designed for an ordinary fault-free asynchronous network, to run on such networks.
KeywordsInternal State Transformation Rule Communication Link Global State State Relay
Unable to display preview. Download preview PDF.
- [cr]1994Springer-VerlagE. Chang and R. Roberts, An improved algorithm for decentralized extrema-finding in circular configuration of processes, CACM 22,5, 1979, 281–283.Google Scholar
- [FF]R.E. Filman and D.P. Friedman, Coordinated computing: tools and techniques for distributed software, McGraw-Hill 1984.Google Scholar
- [FL]G. N. Frederickson and N. A. Lynch, Electing a leader in a synchronous ring, Journal of the ACM, No. 1, Vol. 34, 1987, 98–115.Google Scholar
- [GS]O. Goldreich and L. Shrira, The effects of link failure on computations in asynchronous rings, Proc. of the ACM Symposium on Principles of Distributed Computing, Calgary, Alberta, Canada, August 1986, 174–185.Google Scholar
- [L]G. LeLann, Distributed systems — towards a formal approach, Information Processing 77, Elsevier Science, New York 1977, 155–160.Google Scholar
- [MZ]M. Molcho and S. Zaks, Distributed algorithms for asynchronous networks with a fault detection ring, M.Sc. Thesis, Department of computer science, Technion, Haifa, Israel, February 1990.Google Scholar