Authors: Vishmitha. E, Madhumitha. M
Abstract: Serverless computing is an emerging paradigm in cloud computing that abstracts infrastructure management from developers and enables fully event-driven execution of applications. Unlike traditional cloud models that rely on continuously running virtual machines, serverless platforms dynamically allocate resources and execute functions only in response to events, thereby improving scalability and resource utilization. This paper presents a comprehensive analysis of serverless computing, focusing on its architectural design, performance characteristics, advantages, and inherent challenges. The core components of serverless architecture, namely Function-as-a-Service (FaaS) and Backend- as-a-Service (BaaS), are examined in detail to illustrate how they support stateless execution, automatic scaling, and rapid application development. A comparative study between serverless computing and traditional virtual machine-based cloud models is conducted with respect to scalability, latency, cost efficiency, and operational complexity. Performance factors such as cold start latency, execution overhead, and throughput under varying workloads are analyzed to highlight the trade-offs involved in adopting serverless systems. Furthermore, this paper discusses critical challenges including security concerns arising from multi-tenancy, vendor lock-in due to provider- specific services, limitations in observability and debugging, and complexities in state management. Finally, the paper outlines future research directions aimed at reducing latency, improving portability, enhancing security mechanisms, and integrating serverless computing with edge and hybrid cloud environments to support next-generation distributed applications.
