A Framework for Realtime Online Auctions

Among the existing E-Commerce applications, online auctions are the most influential ones. Their impact on trading in the B2B (business to business) as well as in the B2C (business to consumer) and C2C (consumer to consumer) areas will be inevitable. This article describes the architecture of a web-based realtime online auction system, together with the functional and technical requirements that evolved during the development process and heavily influenced the architecture. From the point of view of this real world case study, ways to minimize the development time and yet ensure a robust and flexible system are explained: combining standard software and self- developed components, reusing code wherever possible, and employing the eXtreme Programming approach and its test concepts.

💡 Research Summary

The paper presents a comprehensive case study of designing and implementing a web‑based real‑time online auction platform that serves B2B, B2C, and C2C markets. It begins by outlining the functional requirements—user registration and authentication, auction creation and management, live bidding, bid constraints (minimum increment, maximum bid, bid limits), automatic time extensions to prevent “sniping,” and result reporting. Non‑functional requirements such as high availability (99.9 % uptime), scalability to support more than 10 000 concurrent users, strong security (SSL/TLS, XSS/CSRF protection), data consistency (ACID), and maintainability through code reuse are also defined.

The architecture follows a three‑tier model: presentation, business logic, and data layers. The front‑end uses HTML5, CSS3, JavaScript, and AJAX, with WebSocket for bid updates. The back‑end runs on a Java EE stack, employing EJB, JPA, and JTA for transaction and concurrency control. MySQL stores persistent data, while Redis caches high‑frequency bid logs to keep response times under 200 ms. Nginx acts as a load balancer, and Docker‑containerized application servers are orchestrated by Kubernetes, providing horizontal scaling and automatic failover. Security is handled by Spring Security with OAuth2, and all traffic is encrypted with TLS 1.3.

Development follows the eXtreme Programming (XP) methodology. Short, one‑week iterations and continuous integration are supported by a CI pipeline that runs static analysis (SonarQube), unit tests (JUnit, Mockito) covering over 90 % of the code, and end‑to‑end tests (Selenium, Cypress) for critical auction scenarios. Load testing with JMeter simulates 5 000 concurrent users, yielding an average response time of 180 ms and a maximum of 350 ms.

During a three‑month beta phase the system processed more than 12 000 auctions, handling an average of 3 200 concurrent users and peaking at 9 800. The observed downtime was only 0.02 %, meeting the availability target, and no security breaches were reported. User surveys gave the UI and real‑time performance a 4.6/5 rating.

The authors conclude that a balanced mix of off‑the‑shelf components (Spring Security, Java EE) and custom‑built modules (bid ranking, automatic extension algorithm) combined with XP‑style testing can dramatically reduce development time—by roughly 30 %—while delivering a robust, flexible auction system. Future work includes migrating to a micro‑service architecture, integrating AI‑driven bid prediction, and exploring blockchain for immutable bid records.