Analyzing the serverless cloud virtualization requirements

• Main Author: Hong, Colin Fung Heng
• Other Authors: Dmitrii Ustiugov
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Computer and Information Science; Serverless computing; Function-as-a-service (FaaS); FaaS platforms; FaaS experimentation; Resource management
• Online Access: https://hdl.handle.net/10356/174840
• Institution: Nanyang Technological University

The evolution of cloud computing has witnessed a significant shift towards microservices architecture, driven by the need for scalability and agility in application development. While microservices offer numerous advantages, including modularity and language heterogeneity, they also present challenges in terms of resource provisioning. To address these challenges, the serverless or function-as-a-service (FaaS) computing model has emerged, promising reduced provisioning times and simplified resource management. However, current serverless platforms suffer from virtualization overhead due to supporting unnecessary syscalls, leading to inefficient resource utilization. This paper contributes to the ongoing endeavor to design lean hypervisors aimed at reducing virtualization overhead by accurately characterizing user code requirements and optimizing resource allocation. By addressing the inefficiencies in serverless hypervisors, this research aims to reduce over-provisioning and improve cost-effectiveness of serverless computing environments.