Comparison of the FCFS and PS discipline in Redundancy Systems

TL;DR

This paper compares FCFS and PS disciplines in redundancy systems with multiple servers, analyzing stability, latency tail behavior, and expected latency to determine their performance differences.

Contribution

It provides a comparative analysis of FCFS and PS disciplines in redundancy systems, focusing on stability and latency metrics, which was not thoroughly explored before.

Findings

FCFS and PS have different stability conditions.

Latency tail behavior varies between FCFS and PS.

Expected latency differs significantly depending on the discipline.

Abstract

We consider the c.o.c. redundancy system with $N$ parallel servers where incoming jobs are immediately replicated to $d$ servers chosen uniformly at random (without replacement). A job finishes service as soon as the first replica is completed, after which all the remaining replicas are abandoned. We compare the performance of the first-come first-served (FCFS) and processor-sharing (PS) discipline based on the stability condition, the tail behavior of the latency and the expected latency.