An Introductory Tutorial to Concurrency-Related Distributed Recursion

Recursion is a fundamental concept of sequential computing that allows for the design of simple and elegant algorithms. Recursion is also used in both parallel or distributed computing to operate on data structures, mainly by exploiting data independence (independent data being processed concurrently). This paper is a short introduction to recursive algorithms that compute tasks in asynchronous distributed systems where communication is through atomic read/write registers, and any number of processes can commit crash failures. In such a context and differently from sequential and parallel recursion, the conceptual novelty lies in the fact that the aim of the recursion parameter is to allow each participating process to learn the number of processes that it sees as participating to the task computation. Key-words: Asynchrony, Atomic read/write register, Branching time, Concurrency, Distributed algorithm, Concurrent object, Linear time, Participating process, Process crash failure, Recursion, Renaming, Shared memory, Task, Write-snapshot. Une introduction à la récursion répartie liée à la concurrence Résumé : Ce rapport constitue une introduction à la récursion répartie lorsque le paramètre de récursivité est utilisé pour capturer le degré de concurrence. Mots clés : Asynchronisme, Calcul réparti, Concurrence, Faute, Registre lire/écrire atomique,Récusrvité, Tâche. * Instituto de Matemarticas, UNAM, Mexico ** Institut Universitaire de France & IRISA (équipe ASAP commune avec l’Université de Rennes 1 et Inria) c ©IRISA – Campus de Beaulieu – 35042 Rennes Cedex – France – +33 2 99 84 71 00 – www.irisa.fr ha l-0 08 58 19 5, v er si on 1 5 Se p 20 13 2 S. Rajsbaum & M. Raynal