In this lecture, Rodrigo will present the two main system and fault models that have been used for the last four decades: the crash fault model, which captures the effects of faults that cause a process or a machine to crash, and the Byzantine fault model, which captures arbitrary failure modes. He will discuss how to design storage protocols under these two models. The lecture will conclude with a discussion on why we might need to revisit these two models in the future.
Hagit Attiya, Amotz Bar-Noy, and Danny Dolev. 1995. Sharing memory robustly in message-passing systems. J. ACM 42, 1 (Jan. 1995), 124–142.
Dahlia Malkhi and Michael Reiter. 1998. Byzantine quorum systems. Distrib. Comput. 11, 4 (October 1998), 203–213.
Slides and exercises