Serverless Computing

Full automation of IT infrastructure and the delivery of efficient IT operations as billed services have been long-standing goals of the computing industry since at least the 1960s. A newcomer--serverless computing--emerged in the late 2010s with characteristics claimed to be different from those of established IT services, including Infrastructure-as-a-Service (IaaS), Platform-as-a-Service (PaaS), and Software-as-a-Service (SaaS) clouds. Even though serverless computing has gained significant attention in industry and academia over the past five years, there is still no consensus about its unique distinguishing characteristics and precise understanding of how these characteristics differ from classical cloud computing. What is serverless computing, and what are its implications? Market analysts are agreed that serverless computing has strong market potential, with projected compound annual growth rates (CAGRs) varying between 21% and 28% through 20284,25,33,35,49 and a projected market value of $36.8 billion49 by that time. Early adopters are attracted by expected cost reductions (47%), reduced operation effort (34%), and scalability (34%).17 In research, the number of peer-reviewed publications connected to serverless computing has risen steadily since 2017.46 In industry, the term is heavily used in cloud provider advertisements and even in the naming of specific products or services. Yet despite this enthusiasm, there exists no common and precise understanding of what serverless is (and of what it is not). Indeed, existing definitions of serverless computing are largely inconsistent and unspecific, which leads to confusion in the use of not only this term but also related terms such as cloud computing, cloud-native, Container-as-a-Service (CaaS), Platform-as-a-Service (PaaS), Function-as-a-Service (FaaS), and Backend-as-a-Service (BaaS).12 As an extended discussion during a 2021 Dagstuhl Seminar2 and our analysis of existing definitions of serverless computing reveal, current definitions focus on a variety of aspects, from abstractions to practical concerns, from computational to financial, from separation of concerns to how concerns should be enacted, and so on. These definitions do not provide consensus, and they are omissive in essential points or even diverge.