UNAM
Usted está aquí: Inicio / Actividades académicas / Eventos / Conferencias Eventuales / Actividades / Distributed termination detection

Distributed termination detection

Ponente: Michel Raynal
Institución: Université de Rennes, France

Cuándo 30/10/2014
de 13:00 a 14:30
Dónde Auditorio "Alfonso Nápoles Gándara"
Agregar evento al calendario vCal
iCal

This  talk  is  on  the detection  of  the termination  of a  distributed computation. This problem was  posed and solved for the first  time in the early  1980s  independently by E.W. Dijkstra and C.S. Scholten (1980) and  N. Francez (1980).  This is a non-trivial problem. While, in sequential computing,  the  termination of  the only  process  indicates  that the computation has terminated, this is no longer true in distributed computing. Even if we were able  to observe  simultaneously all the processes, observing  all of them passive could not allow us to conclude that the distributed execution has terminated. This  is because  some messages  can still  be in  transit, which  will reactivate their  destination processes when  they arrive,  and these re--activations will, in turn, entail the sending of new messages, etc.

The talk  will presents several models of asynchronous computations and observation/detection  algorithms suited to termination detection in each of them.  It corrssponds to chapter 14 of M. Raynal's following book: Distributed Algorithms for Message-Passing Systems, Springer, 2013

Prof. Michel Raynal is among the top researchers in the world on the topic of distributed algorithms. He is a full professor at the Université de Rennes where he founded the Distributed Algorithms research group in 1984. He has been the principal investigator in numerous related research national and international projects, and he has been invited by more than 25 universities around the world to give lectures on distributed algorithms and distributed computing. He has over 300 academic publications on this topic, and has authored a number of books on related topics. His current research interests include distributed algorithms, distributed computing systems, distributed computability and dependability; and the fundamental principles that underlie the design and the construction of distributed computing systems.

El Prof. MIchel Raynal estará visitándonos del 27 de octubre al 7 de noviembre y estará disponible para quien quiera platicar con él. Mayores informes con rajsbaum@im.unam.mx