A History of the Virtual Synchrony Replication Model

Abstract

Abstract.In October of 2015, we celebrated the 50 th anniversary of SOSP as part of SOSP 25. Peter Denning formed a "history day" steering committee, and invited me to give a short talk on the topic of fault tolerance (and also asked if I could help organize the remainder of the day). This essay is intended as an accompaniment to the video and slides of my talk.My topic here is dominated by two fairly specific questions, central to the way we think about the discipline: Must strong properties bring complexity, poor scalability, high latencies and other significant costs? Can a system support consistency without dictating to its users? The debate surrounding these issues has animated the community at least since the mid 1980's.Any decision to focus dictates a degree of narrowness. For example, Butler Lampson has argued that one cannot have security without reliability, and vice versa [25], and the SOSP 2015 program supports his view. Nonetheless, I won't be discussing security here. I hope that nobody is offended by my omission of this and other important work; just like the other history day speakers, I was required to keep the scope of my talk manageable and focused, and omissions were unavoidable.Introduction.

Self-stabilizing Virtual Synchrony

“…Birman et al were the first to present VS and a series of improvements in the efficiency of ordering protocols [5]. Birman gives a concise account of the evolution of the VS model for SMR in [4].…”

“…Virtual Synchrony (VS) has been proven to be very important in the scope of fault-tolerant distributed systems [4]. The VS property ensures that two or more processors that participate in two consecutive communicating groups should have delivered the same messages.…”

“…The challenges related to VS consist of the need to maintain atomic message delivery in the presence of asynchrony and crash failures. VS facilitates the implementation of a replicated state machine [4] that is more efficient than classical consensus-based implementations that start every multicast round with an agreement on the set of recently live and connected processors. It is also usually easier to implement [4].…”

“…VS facilitates the implementation of a replicated state machine [4] that is more efficient than classical consensus-based implementations that start every multicast round with an agreement on the set of recently live and connected processors. It is also usually easier to implement [4]. To the best of our knowledge, no self-stabilizing virtual synchrony solution exists.…”

High availability, elasticity, and strong consistency for massively parallel scans over relational data

“…Chockler et al [22] provide an overview of the theory of group communication; Defago et al give a survey of the many known protocols and their respective guarantees [28]; and Birman [10] provides a historical account. Notable examples of group communication toolkits are Spread [3], Totem [4], JGroups [30], Isis [12], Isis 2 [11], Transis [31], QuickSilver [59], and Horus [75].…”