FRAPPE: Fast Replication Platform for Elastic Services

TL;DR

FRAPPE is a novel replication platform designed to enhance the elasticity of cloud services by employing speculative execution in a new replicated state machine protocol, enabling continuous operation during reconfigurations and failures.

Contribution

The paper introduces FRAPPE, a new platform with a speculative state machine protocol that improves elasticity and fault tolerance in cloud service replication.

Findings

Enables continuous service during reconfiguration

Uses speculative execution for fault tolerance

Improves elasticity of cloud services

Abstract

Elasticity is critical for today's cloud services, which must be able to quickly adapt to dynamically changing load conditions and resource availability. We introduce FRAPPE, a new consistent replication platform aiming at improving elasticity of the replicated services hosted in clouds or large data centers. In the core of FRAPPE is a novel replicated state machine protocol, which employs speculative executions to ensure continuous operation during the reconfiguration periods as well as in situations where failures prevent the agreement on the next stable configuration from being reached in a timely fashion. We present the FRAPPE's architecture and describe the basic techniques underlying the implementation of our speculative state machine protocol.