State-Aware Performance Reproducibility in Stateful Serverless Applications

Abstract

This research paper investigates performance reproducibility in stateful serverless applications. It examines how state-related factors, including state size and state access patterns, interact with workload characteristics and execution environments to influence performance variability. The study focuses on understanding the sources of performance variance and identifying factors that affect the reproducibility of performance measurements in stateful serverless systems. The findings aim to provide a foundation for more reliable and reproducible performance evaluation of stateful serverless applications.