Benefits of Stabilization versus Rollback in Eventually Consistent Key-Value Stores

TL;DR

This paper compares self-stabilization and rollback methods for handling consistency violations in eventually consistent key-value stores, finding stabilization often outperforms rollback and can significantly speed up programs.

Contribution

It provides an empirical evaluation showing stabilization is generally more effective than rollback for managing consistency violations in distributed key-value stores.

Findings

Self-stabilization outperforms rollback in experiments.

Allowing more CVFs can speed up programs by 2-15 times.

Analysis of factors influencing performance results.

Abstract

In this paper, we evaluate and compare the performance of two approaches, namely self-stabilization and rollback, to handling consistency violation faults (cvf) that occurred when a distributed program is executed on eventually consistent key-value store. We observe that self-stabilization is usually better than rollbacks in our experiments. Moreover, when we aggressively allow more cvf in exchange of eliminating mechanisms for guaranteeing atomicity requirements of actions, we observe the programs in our case studies achieve a speedup between 2--15 times compared with the standard implementation. We also analyze different factors that contribute to the results. Our results and analysis are useful in helping a system designer choose proper design options for their program.