<![CDATA[The development of Moodle e-learning utilizing cloud computing services is increasingly popular where the cloud offers a pay-as-you-go resource rental system. Amazon web service (AWS) offers features and services such as AWS EC2 virtual private servers and AWS RDS services as database services for building Moodle services. However, due to the nature of cloud services, namely “pay-as-you-go†or pay according to resource usage, performance is key. To find out the performance of a number of resources that are built, it is necessary to design the resources to be used, looking at the performance aspect compared to the estimated costs that will be incurred as a consideration for resource selection decisions to build Moodle services. Testing and evaluating performance on multiple EC2 resource clusters and single RDS databases and read-replica databases via 50, 100, and 200 tiered HTTP requests on Moodle activities such as Login, Course, Forum, and Quiz to get results on CPU, RAM, and Throughput usage and Latency using Apache JMeter and System Activity Report (SAR). The system test results show that there is an increasing trend along with the increase in requests from clients to servers. View activities such as Login and Course require faster processing time than Forum and Quiz. To calculate the Performance-to-Cost Ratio, the EC2 t2.small and RDS db.t2.small clusters are superior from the aspect of Throughput to cost ratio of 0.23 requests/second per dollar.]]>
Copyrights © 2022
