FaaSOnto: A Semantic Model for Enabling Function-as-a-Service Platform Selection

Serverless computing shifts the responsibilities of provisioning and managing cloud infrastructure resources from developers to cloud service providers, allowing developers to focus solely on their applications. Function-as-a-Service (FaaS) is a serverless computing approach that enables developers to develop their applications as event-driven functions. There are many FaaS platforms available through public cloud providers or open-source distributions. Understanding the differences in these platforms and keeping up to date with their latest developments is challenging. Hence, it is necessary to systematically model the information about FaaS Platforms to allow practitioners to select the platform most suited for realizing their use cases. This paper presents the FaaSOnto ontology, a semantic model that represents the characteristics of FaaS platforms. We developed the ontology systematically following the NeOn methodology. We fully implemented the ontology using OWL2 and created a knowledge base with information about ten different FaaS platforms. The knowledge base is semi-automatically populated. On top of the knowledge base, we developed a minimal decision support system to enable the sorting and filtering of FaaS platforms based on their characteristics to facilitate an interactive platform selection process.