The 2025 Unified Tertiary Matriculation Examination (UTME), conducted by the Joint Admissions and Matriculation Board (JAMB), was unfortunately marred by significant technical difficulties which led to an unprecedented acknowledgement by JAMB and an emotional response from its Registrar, Professor Is-haq Oloyede, who was moved to tears. These unexpected challenges affected thousands of candidates nationwide, causing frustration and widespread public concern. The UTME is a critical step for many students, and this disruption can significantly impact their educational futures. As a Security Operations Centre analyst and a Cloud Engineering student, I felt compelled to investigate the situation to understand the technical reasons behind it.

Relying on fundamental principles of modern digital infrastructure, system reliability, and incident response, I will examine the likely factors that contributed to these system failures and suggest ideas for future implementations that could help create a more robust and adaptable technical architecture.

The 2025 glitch Is traceable to some changes JAMB made in order to improve the integrity and efficiency of the exam. The changes introduced were to enhance result analysis integrity via a shift from count-based analysis to source-based analysis and implement a full-scale shuffling of question-and-answer options and to improve the performance and responsiveness of the platform.

These changes were meant to reduce exam malpractice, improve auditability, and minimise latency and system crashes during mass CBT sessions. However, JAMB implemented these changes on a server-cluster-per-zone basis, rather than uniformly across all clusters, creating inconsistencies in performance and result output between zones.

Technical analysis:

The updates by JAMB applied to specific zones, like Kaduna but not Lagos, suggest that JAMB uses multiple regional server clusters. A critical factor appears to be the potential of these regional clusters to operate with some independence in technical management, including applying updates.

Applying updates to some clusters but not others creates a fragmented technical environment, known as “patch drift”. This inconsistency can lead to several problems like unequal exam experience, data inconsistency, increased security risks, operational complexity, and delays.

The lack of a synchronised and uniform deployment strategy appears to be a primary technical factor behind the widespread issues.

No doubt, the technical plans JAMB put in place for 2025’s examination lacked a proper implementation strategy. To avert a recurrence in subsequent examinations, JAMB needs a centralised, automated, and validated system rollout pipeline that would ensure uniform change. Specifically, JAMB can do the following:

1. Use a centralised control plane for patch distribution: to define and automatically push updates, configurations, and scripts to all server clusters simultaneously. This ensures every part runs the same approved version, eliminating patch drift.

2. Use automated update rollout using Infrastructure as Code (IaC) to manage infrastructure through code (e.g., Terraform, Ansible): Automated deployment pipelines detect changes, validate them against tests, and push to all zones. Built-in safeguards like retries, rollbacks, and approval gates minimise human error.

3. Conduct canary testing and pre-exam validation: Before full deployment, test significant changes in a small, isolated, or simulated environment. “Canary testing” observes system behaviour under realistic conditions (response times, shuffling, result consistency) without impacting most candidates, allowing for adjustments.

4. Post-deployment validation & observability: Continuously monitor the system using tools for metrics collection and log analysis. Automated checks compare expected vs. actual outcomes (e.g., shuffled results, biometric matches) and alert technical teams to discrepancies quickly. This observability enables proactive problem-solving.

5. High Availability (HA) through backup clusters and automated failover: Implement standby server clusters in each zone that are kept up-to-date. Automated failover mechanisms redirect traffic to standby clusters if primary ones fail, minimising disruption. Backups must be patched simultaneously with primary servers for consistency.

6. Comprehensive incident response plan: Develop and regularly test a detailed plan for technical teams to follow during incidents. This includes early detection of problems, communicating internally and externally, mitigating impact, and restoring operations. A practised plan ensures a coordinated and effective response, minimising downtime and confusion.

In conclusion, the 2025 JAMB challenge highlights the importance of having a resilient, scalable, and well-orchestrated infrastructure, especially for a system that serves millions of users. Leveraging tools like Infrastructure as Code, carrying out canary testing on new updates before deploying, and developing an incident response plan can make high-demand platforms like JAMB’s system become more reliable, secure, and adaptable.

This analysis is a constructive contribution to the conversation on how modern technology and engineering principles can support critical national-scale digital services in Nigeria. Learning from these experiences is key to building stronger, smarter public technology infrastructures for everyone’s benefit.

If implemented, we are certain that there would not be outrage next year, and we would be spared tears on national television. (BusinessDay)