A Comparative Study of Load Balancing Methods in Cloud-Based Village Information Systems

Authors:

Abstract
Internet usage has been steadly increasing, with many needs or information that must be sought. The problem that is often experienced when many visitors access a website is that the server acting as a web server is unable to handle requests so services cannot be processed. This can be handled using the load balancing method using HAProxy. A web-based application certainly requires an adequate server computer so that the application can run properly. Cloud computing has emerged as a new trend that promotes efficient resource utilization and optimal server infrastructure utilization. Presently, numerous cloud computing platforms are offered, one of which is Amazon Web Services (AWS). In this research, several stages were carried out, starting from literature study, software design, testing stages and analysis of the results of testing data, drawing conclusions, and finishing. In designing cloud infrastructure, researchers created three services on AWS including EC2 (Elastic Cloud Compute), RDS (Relational Database Service), and VPC (Virtual Private Cloud) services. The village information system's load balancing design successfully applied to cloud technology, Amazon Web Service, and successfully implemented HAProxy as the load balancer and OpenSID as the village information system's website. the HA proxy employs various algorithms for a website, including round robin, least connections, source IP Hash, and Uniform Resource Identifier (URI). By using the parameters of standard deviation, throughput, average bytes and error in performance tests consisting of load tests and soak tests in this research, it has been determined that the URI algorithm is the most suitable for OpenSID.