Metodología para la recolección y curación de un dataset de aplicaciones serverless (desarrolladas con Serverless Framework y AWS Lambda) de código abierto en GitHub para su caracterización

Abstract

Function-as-a-Service (FaaS) is at the core of serverless computing, enabling developers to easily deploy applications without managing computing resources. With an Infrastructure-as- Code (IaC) approach, frameworks like the Serverless Framework use YAML configurations to define and deploy APIs, tasks, workflows, and event-driven applications on cloud providers, promoting zero-friction development. As with any rapidly evolving ecosystem, there is a need for updated insights into how these tools are used in real-world projects. Building on the methodology established by the Wonderless dataset for serverless computing (and applying multiple new filtering steps), OpenLambdaVerse addresses this gap by creating a dataset of current GitHub repositories that use the Serverless Framework in applications that contain one or more AWS Lambda functions. We then analyze and characterize this dataset to get an understanding of the state-of-the-art in serverless architectures based on this stack. Through this analysis we gain important insights on the size and complexity of current applications, which languages and runtimes they employ, how are the functions triggered, the maturity of the projects, and their security practices (or lack of). OpenLambdaVerse thus offers a valuable, up-to-date resource for both practitioners and researchers that seek to better understand evolving serverless workloads. Index Terms: serverless, serverless computing, function-as-aservice, cloud computing, characterization, repository mining.