Unveiling Overlooked Performance Variance in Serverless Computing

TL;DR

This paper reveals significant, often overlooked performance variability in serverless functions, demonstrating that ignoring this variance can impact research reproducibility and reliability.

Contribution

It uncovers the extent of performance variance in serverless computing and highlights the community's lack of awareness and inadequate measurement practices.

Findings

Performance can differ by up to 338.76% across runs

61.11% of functions produce unreliable results

Most studies use too few repetitions to measure variance

Abstract

Serverless computing is an emerging cloud computing paradigm for developing applications at the function level, known as serverless functions. Due to the highly dynamic execution environment, multiple identical runs of the same serverless function can yield different performance, specifically in terms of end-to-end response latency. However, surprisingly, our analysis of serverless computing-related papers published in top-tier conferences highlights that the research community lacks awareness of the performance variance problem, with only 38.38% of these papers employing multiple runs for quantifying it. To further investigate, we analyze the performance of 72 serverless functions collected from these papers. Our findings reveal that the performance of these serverless functions can differ by up to 338.76% (44.28% on average) across different runs. Moreover, 61.11% of these functions produce unreliable performance results, with a low number of repetitions commonly employed in the serverless computing literature. Our study highlights a lack of awareness in the serverless computing community regarding the well-known performance variance problem in software engineering. The empirical results illustrate the substantial magnitude of this variance, emphasizing that ignoring the variance can affect research reproducibility and result reliability.