A review on serverless architectures - function as a service (FaaS) in cloud computing

Abstract

Emergence of cloud computing as the inevitable IT computing paradigm, the perception of the compute reference model and building of services has evolved into new dimensions. Serverless computing is an execution model in which the cloud service provider dynamically manages the allocation of compute resources of the server. The consumer is billed for the actual volume of resources consumed by them, instead paying for the pre-purchased units of compute capacity. This model evolved as a way to achieve optimum cost, minimum configuration overheads, and increases the application's ability to scale in the cloud. The prospective of the serverless compute model is well conceived by the major cloud service providers and reflected in the adoption of serverless computing paradigm. This review paper presents a comprehensive study on serverless computing architecture and also extends an experimentation of the working principle of serverless computing reference model adapted by AWS Lambda. The various research avenues in serverless computing are identified and presented.

Mobile Health in Remote Patient Monitoring for Chronic Diseases: Principles, Trends, and Challenges

“…This architecture is associated with terms such as function-as-a-service (FaaS), which addressed various needs that were not handled by PaaS and SaaS. A comprehensive survey about FaaS is available in [ 56 ].…”

A Traffic Analysis on Serverless Computing Based on the Example of a File Upload Stream on AWS Lambda

“…The growing need for a shared network of nodes, fast response rates of services, the resulting agility of resource provisioning, and minimised management during run-time have been targets of hyper-scaling microservice platforms in recent years [2,10]. This refers to the pooling and utilising of resources including operating systems, runtime environments and hardware on a pay-on-demand basis [10]. The main concept behind this is called microservisation.…”

“…The approach frees the user from having to maintain a server, including configuration and management of VMs, as resource management is fully provided by the platform in an automated and scalable way. Billing is promised to be limited to the actual volume of resources consumed [10]. The difference between Platform as a Service and Serverless is that the latter not only scales on-demand but also runs on-demand, as providers enable instances to go cold.…”

“…The difference between Platform as a Service and Serverless is that the latter not only scales on-demand but also runs on-demand, as providers enable instances to go cold. This shifts the runtime concerns to the provider and leaves the developer needing only to consider the programming model [10,12]. Balancing the right size and number of services should be the only architectural concept the developer has to think about.…”

“…As AWS Lambda is most-used and associated with the most popular Serverless platform, it is taken as the base platform for research [2,10]. Most studies are based on the configuration and performance of containers and VMs in general.…”