A Cloud-Native Event Ticketing Platform Leveraging Automated Horizontal Scaling and Continuous Deployment through AWS CodePipeline

Abstract: Online event ticketing services routinely experience extreme, short-lived demand surges when high-profile events open for sale, frequently overwhelming statically provisioned infrastructure and producing failed transactions, duplicate seat allocation, and revenue loss. This work presents the design, implementation, and empirical evaluation of a cloud-native ticketing platform engineered to absorb such volatility through elastic horizontal scaling and a fully automated delivery pipeline. The backend services are implemented in Python together with a Node.js-based presentation tier and are deployed across an Amazon Web Services (AWS) environment in which an Application Load Balancer distributes traffic over an Auto Scaling Group whose capacity is governed by real-time CloudWatch utilization signals. A distributed locking strategy backed by an in-memory cache is introduced to enforce seat-level consistency under concurrent booking attempts. Continuous integration and continuous deployment are realized through AWS CodePipeline, CodeBuild, and CodeDeploy using a blue-green release model with automatic rollback. Experimental load testing demonstrates that the proposed architecture sustains sub-300 ms average response times at eight thousand concurrent users while a comparable monolithic baseline degrades beyond three seconds. The system further reduces deployment lead time by roughly 78% and eliminates double-booking under contention. The principal contributions are an elasticity-aware ticketing reference architecture, a consistency-preserving concurrency mechanism, and a reproducible automated deployment workflow.

Keywords: Cloud computing; auto scaling; continuous deployment; event ticketing; AWS CodePipeline; load balancing; concurrency control; DevOps