Performance Analysis of CI/CD Pipelines for Terraform-based Docker Deployments

Mahale, Nirmal Sharadchandra (2025) Performance Analysis of CI/CD Pipelines for Terraform-based Docker Deployments. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (890kB) | Preview

Abstract

Cloud-native software is heavily dependent on Continuous Integration and Continuous Delivery (CI/CD) pipelines to facilitate the process of automated, scalable and efficient deployment of applications. Nevertheless, the organizations lack empirical advice regarding the CI/CD platform is most effective with the Infrastructure-as-Code and containerised workloads, especially when implemented using Terraform in cloud providers. This gap is addressed in this report by comparing three popular CI/CD tools – GitHub Actions, Jenkins and AWS CodePipeline, performing the running of Dockerized Django applications Dockerized using Terraform on AWS. In order to maintain a consistent evaluation, a bespoke Metrics API and real-time Dashboard was developed to assemble five new cloud-specific performance measurements, namely: Layer Cache Efficiency (LCE), Pipeline Recovery Time (PRT), Secrets Management Overhead (SMO), Dynamic Environment Time (DEPT), and Container Layer Build Consistency (CLBC). The results demonstrate that GitHub Actions is the most ideal cloud-native and has excellent caching performance, stable image building, and expeditious recovery performance. Jenkins had poor caching and environmental variability with self-hosted runners as well as being highly configurable. The AWS CodePipeline was highly reproducible, lacks have high provisioning overhead as it is a multi-layered, orchestrated deployment. These results validate the existence of large differences in cloud- native CI/CD behaviors even with the same workload, and that the conventional metrics cannot be used to reflect more profound operational features. The research offers a reproducible benchmarking model and offers viable implications to DevOps teams choosing CI/CD applications to deploy cloud resources using Terraform. This finding will provide a base of future studies that will include multi-cloud comparisons, microservices-scale testing, and incorporating DevSecOps performance metrics.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Jaswal, Shivani UNSPECIFIED
Subjects:	T Technology > T Technology (General) > Information Technology > Cloud computing Q Science > QA Mathematics > Computer software > Communications software T Technology > T Technology (General) > Information Technology > Computer software > Communications software
Divisions:	School of Computing > Master of Science in Cloud Computing
Depositing User:	Ciara O'Brien
Date Deposited:	01 Sep 2026 09:13
Last Modified:	01 Sep 2026 09:13
URI:	https://norma.ncirl.ie/id/eprint/9720

Actions (login required)

View Item