Guidelines for web server overload management for e-learning systems. Case of higher institution of learning in Uganda

Abstract

Globally, e-learning has been adopted as an alternative learning method due to the outbreak of the COVID-19 pandemic in a bid to curb the spread of the virus. In Uganda, the Ministry of Education and Sports (MoES) together with Uganda National Council of Higher Education (UNCHE) issued a directive for all education institutions to adopt e-learning to enable continuity of learning. Because of this directive, customized e-learning systems used by the different institutions of higher learning in Uganda became a better alternative of learning which led to massive uptake of e-learning. E-learning systems are therefore faced with web server overload due to concurrency access especially during peak time leading to poor performance. This affected students in many ways for instance, postponing of examinations, quizzes, tests among others which led to a shift to other platforms such as zoom, Microsoft teams, google classroom among others due to frustration with the poor performance in terms of slow response time. The aim of the research study was to develop guidelines to improve on the performance of e-learning systems that are used in higher institutions of learning in Uganda. The study adapted the design science research methodology with mixed research methods. The first objective was achieved by conducting interviews which led to the discovery of web server overload causes for the e-learning systems. The second objective was also achieved by conducting interviews with the purposively selected participants which led to the determination of web server overload management strategies deployed by the institutions. These strategies contributed to the development of web server overload management guidelines which was the third objective. The developed web server overload management guidelines were later validated based on the Likert scale to ascertain their relevance and anticipated outcome or implications on the e-learning systems when implemented. The study developed web server overload management guidelines that can be adopted to improve on the performance of the e-learning systems used at higher institution of learning in Uganda. These guidelines include: the use of least connection load balancing method; configuring a backup or secondary load balancer; hosting e-learning systems on dedicated web servers; upgrading of the database server from MySQL to PostgreSQL Database Management System; clustering of PostgreSQL database into “read only” and “read and write” databases; optimization of the databases through indexing and caching of information using Redis in-memory software. These guidelines can be used in other systems that face web server overload coupled with high concurrency access during peak time.