DEVELOPMENT OF A BACKEND SYSTEM IN THE 2024 ELECTRONIC ELECTION RECAPITULATION SYSTEM APPLICATION

• Main Author: Dimas Baskara, Gregorius
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/75269
• Institution: Institut Teknologi Bandung

Indonesia is the third largest democracy in the world which will hold simultaneous elections and local elections in 2024. Based on a statement by the General Election Commission of the Republic of Indonesia, 190.022.169 voters will vote in the event. Elections are carried out conventionally where voters vote at TPS. The voting process is then followed by the vote counting and recapitulation process. Vote recapitulation is carried out in stages from the sub-district, regency/city, provincial and national levels. With a very large number of voters and TPS, this manual recapitulation process raises several problems. As an effort to overcome this problem, the Sirekap mobile application, which was used in the 2020 local elections, has been re-developed for use at the 2024 democratic party. An important part of the Sirekap application is a scalable backend system. This final project aims to build a backend system that is able to serve every request well. Aspects that are considered in this Final Project are database schemes, flow and implementation at the application level, and application deployment architecture to support application performance in dealing with large numbers of requests. In the aspect of deployment architecture, Kubernetes infrastructure and serverless computing will also be compared in this Final Project. The solution that was successfully developed is an application using the Golang programming language connected to the PostgreSQL database and deployed both server-based and serverless Kubernetes on Alibaba Cloud and Google Cloud. The development of the backend system is based on API requirements by the frontend system and follows the database scheme and flow of the Sirekap 2020 backend system. After the system has been built, load testing is carried out to test the system's ability to deal with many requests at the same time. The results of load testing show that the system has been able to serve peak requests similar to what will be encountered on election day or regional elections in 2024 with variations in results for each different development environment.