Server Cluster Patch Failure Responsible For UTME Retake, Expert Says
…No Regional Bias, Says Chidoka
A critical technical glitch in the 2025 Unified Tertiary Matriculation Examination (UTME) has forced the Joint Admissions and Matriculation Board (JAMB) to reschedule exams for 379,997 candidates from today, May 16, to May 18, due to a failed software patch in its server clusters.
THE WHISTLER reported that the failure affected 157 exam centres in Lagos and the Southeast, disrupting the examination process for a significant portion of the 1,955,069 candidates.
“We realised from their complaints and reports that a system patch was necessary to support the whole shuffling – the whole question and answer shuffling – and it was meant to be fully deployed across the two server clusters that support KAD and LAG, but there was an omission,” Engineer James Nnanyelugo of Educare’s Technical Team explained on a Twitter space.
Server clusters are groups of interconnected computers that work together to process large-scale data, such as delivering UTME content to Computer-Based Test (CBT) centres. JAMB uses two clusters—KAD and LAG—to manage exam content delivery, relying on load balancing to handle 1.9 million candidates’ data.
The KAD cluster serves most Northern states, excluding six (Kano, Katsina, Jigawa, Niger, Kogi, and FCT), and includes six South-South states. The LAG cluster covers all Southern states (except those South-South states), the six excluded Northern states, and the rest of the Southwest (Ogun, Oyo, Osun, Ondo, and Ekiti). The Southeast (92 centres, 173,387 candidates) and Lagos (65 centres, 206,610 candidates), both under the LAG cluster, were hit hardest due to the uneven patch deployment.
Nnanyelugo noted, “They sent that patch. They had their quality assurance test on the centres that were in the patch, taking random tests, samples, and exams from the centres that they selected.
“According to the reports, everything passed before they gave the go-ahead order to the vendors to publish, that is, to synchronise with the base stations.”
It was noted that 157 LAG centres missed the patch, disrupting sessions 1–16, each with 250 candidates.
“JAMB, following the results, noticed that the patch was not doing what it should do, that is, shuffling the results, and they called the attention of the handlers, who made a bug fix. I will call it a bug fix now because it’s supposed to be a system that has been tested, trusted, and pushed.
“So they made a bug fix and sent it out. Now, this will be the second time that they are sending it out, but this time around, the exam was already going on. And that was in the seventeenth session.
“So, from everybody who wrote the exam within the seventeenth session, the results were okay all the way up. I also got to understand that an attempt was made to patch the previous 16,” Engr. James added.
The failure to synchronise the patch across the LAG cluster’s centres led to the decision to reschedule exams for affected candidates.
Addressing concerns about regional bias, particularly the heavy impact on the Southeast, the former Minister of Aviation and founder of The Athena Centre, Osita Chidoka, who participated in the review, stated, “I was particular in the conversation because when we narrowed down to the Southeast, I then knew that we had a problem because that meant that a zone, you know, is almost like being targeted. But I saw the contractor. I saw the consultant. I heard their story as narrated by Nnayelugo.
“From all the evidence before us and from my queries at the review, I do not think there was a deliberate effort to undermine the Southeast. I don’t think it was any deliberate effort. It was a technical fault. It was a mistake, and it speaks to the wider issue of IT professionals in Nigeria.”
Chidoka noted the rigorous review process, including manual remarking of scripts from other states, which took 30 people a day, and collaboration with stakeholders.
He added, “What is most important for me is that system failure is a part of every technical system. So we should not make a fetish of technology that it cannot fail. It can fail. But what is important and what is happening today is that when it fails, you must come clean with what happened so that lessons can be learnt, systems can be upgraded, and the stakeholders can see that you can have oversight into what happened.”
Server Cluster Patch Failure Responsible For UTME Retake, Expert Says is first published on The Whistler Newspaper
