Issue No.10 - October (1985 vol.11)
An efficient decentralized algorithm for synchronized termination of a distributed computation is presented. It is assumed that distributed processes are connected via unidirectional channels into a strongly connected network, in which no central controller exists. The number of processes and the network configuration are not known a priori. The number of steps required to terminate distributed computation after all processes met their local termination conditions is proportional to the diameter D of the network (D + 1 steps).
termination detection, Decentralized algorithms, distributed programming, distributed systems, networks
null Yuan Shi, N.S. Prywes, "Synchronized Distributed Termination", IEEE Transactions on Software Engineering, vol.11, no. 10, pp. 1136-1140, October 1985, doi:10.1109/TSE.1985.231861