Analysis of the Matrix Event Graph Replicated Data Type

TL;DR

This paper introduces the Matrix Event Graph (MEG), a replicated data type for decentralized systems that ensures strong eventual consistency and scalability while tolerating Byzantine faults, with implications for secure messaging and distributed applications.

Contribution

The paper formally analyzes MEG as a Conflict-Free Replicated Data Type, demonstrating its properties for consistency, fault tolerance, and scalability in decentralized environments.

Findings

MEG provides Strong Eventual Consistency (SEC).

MEG is resilient to Byzantine faults with n > f participants.

The width of MEG evolves optimally over time, indicating good scalability.

Abstract

Matrix is a new kind of decentralized, topic-based publish-subscribe middleware for communication and data storage that is getting popular particularly as a basis for secure instant messaging. In comparison to traditional decentralized communication systems, Matrix replaces pure message passing with a replicated data structure. This data structure, which we extract and call the Matrix Event Graph (MEG), depicts the causal history of messages. We show that this MEG represents an interesting and important replicated data type for general decentralized applications that are based on causal histories of publish-subscribe events: we show that a MEG possesses strong properties with respect to consistency, byzantine attackers, and scalability. First, we show that the MEG provides Strong Eventual Consistency (SEC), and that it is available under partition, by proving that the MEG is a Conflict-Free Replicated Data Type for causal histories. While strong consistency is impossible here as shown by the famous CAP theorem, SEC is among the best known achievable trade-offs. Second, we discuss the implications of byzantine attackers on the data type's properties. We note that the MEG, as it does not strive for consensus, can cope with $n > f$ environments with $n$ total participants of which $f$ show byzantine faults. Furthermore, we analyze scalability: Using Markov chains we study the width of the MEG, defined as the number of forward extremities, over time and observe an almost optimal evolution. We conjecture that this property is inherent to the underlying spatially inhomogeneous random walk.