Brief Announcement: Self-stabilizing Virtual Synchrony

Shlomi Dolev (1) , Chryssis Georgiou (2) , Ioannis Marcoullis (2) , Elad M. Schiller (3)

1 BGU - Ben-Gurion University of the Negev
2 UCY - University of Cyprus [Nicosia]
3 Chalmers University of Technology [Göteborg]

Systems satisfying the Virtual Synchrony (VS) [2] property provide message multicast and group membership services in which all system events, group membership changes, and incoming messages, are delivered in the same order. VS is an important abstraction, proven to be extremely useful when implemented over asynchronous, typically large-scale, message-passing distributed systems, as it simplifies the design of distributed applications, e.g., State Machine Replication (SMR). The VS property ensures that two or more processors that participate in two consecutive communicating groups should have delivered the same messages. Self-stabilizing systems [1,3] can tolerate transient faults that drive the system to an unpredicted arbitrary configuration. Such sys- tems automatically regain consistency from any such configuration, and then produce the desired system behavior ensuring it for a practically infinite number of successive steps, e.g., 264 steps. We present the first, to our knowledge, self-stabilizing virtual synchrony algorithm.

Domains: Distributed, Parallel, and Cluster Computing [cs.DC]

Shlomi Dolev, Chryssis Georgiou, Ioannis Marcoullis, Elad M. Schiller. Brief Announcement: Self-stabilizing Virtual Synchrony. DISC 2015, Toshimitsu Masuzawa; Koichi Wada, Oct 2015, Tokyo, Japan. ⟨hal-01207862⟩