Timestamps for Partial Replication

TL;DR

This paper presents a new algorithm for maintaining causal consistency in distributed shared memory systems with partial replication, addressing scalability issues of full replication methods.

Contribution

It introduces a vector timestamp-based algorithm tailored for partial replication, linking it to existing full replication approaches and analyzing its timestamp size costs.

Findings

The algorithm effectively tracks causal dependencies in partial replication.

Timestamp size varies based on data sharing among replicas.

The approach generalizes previous full replication methods.

Abstract

Maintaining causal consistency in distributed shared memory systems using vector timestamps has received a lot of attention from both theoretical and practical prospective. However, most of the previous literature focuses on full replication where each data is stored in all replicas, which may not be scalable due to the increasing amount of data. In this report, we investigate how to achieve causal consistency in partial replicated systems, where each replica may store different set of data. We propose an algorithm that tracks causal dependencies via vector timestamp in client-server model for partial replication. The cost of our algorithm in terms of timestamps size varies as a function of the manner in which the replicas share data, and the set of replicas accessed by each client. We also establish a connection between our algorithm with the previous work on full replication.