The adoption of Learning Management Systems (LMS) has significantly accelerated, particularly following the COVID-19 pandemic, as institutions seek scalable solutions for remote and blended learning. Open-source platforms like Moodle have become pivotal in this regard, offering cost-effective, customizable environments for teaching. However, institutions in Africa often encounter numerous challenges in LMS deployment, including infrastructure limitations and scalability constraints. This paper is a practitioner case study involving the implementation of High-Performance Computing (HPC) to e-learning infrastructures. Based on quantitative performance monitoring of Moodle using Cacti, the study covers six federal universities in Nigeria. Data were collected before and after the migration of a centralized single server system, in which sixteen universities were sharing a common environment, to a distributed, cloud-based system with independent virtual servers per university. The results show improvements in performance: the average latency was reduced by 420 to 230 (around 45 percent) and the page load time was reduced to 2.1s (previously 3.8s) at peak load, the error rates were also lowered (4.3 percent to 1.2 percent) and the overall availability did not change (99.12 percent to 99.34 percent). Importantly, the distributed architecture enabled individual universities to upgrade Moodle (4.1 to 5.0) and PHP without interfering with the services of other institutions, which enabled the self-paced learning capabilities and the addition of new plugins. Findings indicate increase in scalability, reliability, and institutional autonomy. The paper concludes with a set of recommendations to strengthen sustainability and capacity building towards the adoption of e-learning in African environments that are resource-constrained.