Monitoring Resource Usage of Digital Learning Platforms for Online and Onsite Learning

Abstract

Technology development brings numerous benefits and transforms the landscape of education. As digital learning has become popular in Taiwan, various digital platforms have been developed. As online services on such digital platforms are diversified according to the increase in the number of digital platforms, the competitiveness of professionals in related industries is demanded more than before. Thus, it is necessary to estimate and decide on an optimal load of resource usage. In this study, we analyzed the data from TronClass of University A in Taiwan to determine the maximum load for resource usage. Data were collected from September 13 to November 7, 2021. The period for a comparative analysis was defined as the online learning period (from September 13 to October 10) and the onsite teaching period (from October 11 to November 7). During the online and onsite learning periods, the highest average CPU usage rate of TronClass was observed on Thursdays, whereas the lowest usage was on Mondays. The highest average CPU and memory usage of the web and database servers was observed also on Thursdays. 1143 and 839 users accessed the Web, while 609 and 473 used mobile devices, and the peak numbers of concurrent web users were 5284 and 2076, while those of mobile device users were 4748 and 2933 during the online and onsite learning periods, respectively. The CPU usage rate of the web server during the online teaching period was 13% higher than that during the onsite teaching period. The CPU usage rate of the database server during the online teaching period was 2−3% higher than that during the onsite teaching period.