A Production-Grade Blueprint for Containerized Web Applications Deployment: Automated Infrastructure and CI/CD Orchestration on AWS

The accelerating adoption of cloud-native software delivery has fundamentally redefined the operational expectations placed on modern web platforms. Enterprises are increasingly required to provision infrastructure on demand, release software continuously, and recover from failures automatically — imperatives that traditional, manually managed deployment architectures are structurally incapable of satisfying. This paper presents the comprehensive design, architecture, and empirical validation of the the proposed car booking system system — a cloud-native, containerized digital car booking platform deployed entirely on Amazon Web Services (AWS) using an integrated DevOps methodology. The proposed architecture addresses the well-documented deficiencies of conventional deployment models, specifically poor horizontal scalability, manual and error-prone release management, infrastructure configuration drift, and the absence of automated fault-recovery mechanisms. The system containerizes the Java-based application using Docker and orchestrates workloads through AWS Elastic Kubernetes Service (EKS), enabling horizontal pod autoscaling and self-healing pod rescheduling. A fully automated Continuous Integration and Continuous Deployment (CI/CD) pipeline orchestrated by Jenkins and integrated with GitHub enforces rapid, repeatable, and auditable release cycles. All AWS infrastructure — including the Virtual Private Cloud (VPC) topology, EC2 node groups, and RDS database instances — is provisioned and managed through Terraform, adhering to Infrastructure-as-Code (IaC) principles. The deployment is hosted within a multi-tier VPC architecture spanning two availability zones, incorporating public and private subnets, NAT gateways, Internet gateways, and an Application Load Balancer (ALB) for traffic distribution and fault isolation. Empirical results demonstrate a mean CI/CD pipeline duration of 7.2 minutes, dynamic horizontal scaling from 2 to 8 pods within 90 seconds of elevated load, zero-downtime rolling deployments across all tested release cycles, and automated pod recovery within 18 seconds of simulated failure. These findings establish the proposed architecture as a reproducible, production-grade reference model for DevOps-driven deployment in high-availability web application domains.