Performance Analysis of Various Server Hosting Techniques
Prerna Jain, Yogesh Munjal, Jatin Gera, Pooja Gupta
Abstract
Cloud Computing is growing exponentially and provides various server hosting services like Virtual Machines, Docker contain- ers and Serverless Architectures. The traditional Virtual Machines (VMs) have long-running power with computation power that appears on request and vanishes after use, right away. Containers are created on top of an operating system using docker; which also automates the deployment of applications on them. Serverless applications are event-driven cloud-based systems where de- velopment of application depends completely on combining the client side’s logic, 3rd parties’ services and cloud-hosted remote procedure calls. The Serverless Container integrates docker containers with serverless applications. Amazon Web Services(AWS) offers ECS with Fargate which hosts docker containers as serverless applications. Serverless Container-aware ARchitecture(SCAR) framework[1] provides the same functionality of hosting docker containers over AWS Lambda. Our work focuses on the perfor- mance analysis of various compute services mentioned above. Based on the results we have also offered recommendations on picking from various server hosting techniques.