Parallel Serverless Workloads in OpenWhisk on Kubernetes

Sharma, Neeti (2022) Parallel Serverless Workloads in OpenWhisk on Kubernetes. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (2MB) | Preview
Preview	PDF (Configuration manual) Download (720kB) | Preview

Abstract

Recent advances in virtualisation technology have resulted in the widespread use of Serverless computing, which provides the ability to programme and administer the cloud in an autoscaling, pay-as-you-go fashion. The platform handles the invocation, execution, and scaling of a function thereby, managing the complete lifecycle of a function. The most famous offering of Serverless computing is Function as a Service (FaaS). Though FaaS provides ease of use, and flexibility it also possesses significant overheads such as vendor-lock-ins, complex pricing models, functional restrictions, and security. In this paper, an open-source serverless framework has been proposed which uses Apache OpenWhisk as FaaS on a single master multi-worker Kubernetes cluster deployed on private cloud NCI OpenStack. The results show that OpenWhisk deployed on Kubernetes works seamlessly on OpenStack with the proposed framework and performs better when the compute intensive workloads are parallelised for C++, Java and Go. The computed results not only demystify the feasibility of OpenWhisk on Kubernetes but also checks the resource utilisation and performance of the system for parallelised workloads.

Item Type:	Thesis (Masters)
Subjects:	Q Science > QA Mathematics > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Electronic computers. Computer science T Technology > T Technology (General) > Information Technology > Cloud computing
Divisions:	School of Computing > Master of Science in Cloud Computing
Depositing User:	Tamara Malone
Date Deposited:	16 Dec 2022 09:35
Last Modified:	16 Dec 2022 09:35
URI:	https://norma.ncirl.ie/id/eprint/5985

Actions (login required)

View Item