Pengujian Secara Fungsional, Struktural, dan Performa pada KMS Desa Digital Indonesia

Abstract

Pengujian merupakan tahapan penting dalam memastikan kualitas dan kesiapan sistem sebelum digunakan secara luas. Pada penelitian ini, pengujian diterapkan pada Knowledge Management System Desa Digital Indonesia (KMS DDI) dengan tiga pendekatan, yaitu black-box, white-box, dan performance, mengikuti Software Testing Life Cycle (STLC). Black-box testing dilakukan menggunakan Katalon Studio melalui kombinasi manual dan automation testing dengan teknik error guessing. White-box testing menggunakan teknik basis path testing dengan framework Jest. Performance testing dilakukan menggunakan Apache JMeter dengan skenario load testing 1–500 virtual users. Hasil black-box dari 312 skenario pengujian menghasilkan 51 error dengan success rate 84%, didominasi functional error. Pada white-box testing, ditemukan lima error dengan success rate 91%. Pada performance testing, sebagian besar endpoint menunjukkan level Apdex Good dan Excellent, namun masih terdapat temuan seperti error dengan pesan “connection reset”, dan request lambat pada beberapa endpoint dan request awal. Seluruh error yang ditemukan kemudian dianalisis untuk diberikan rekomendasi perbaikan agar dapat meningkatkan kualitas sistem.

Testing is a critical stage in ensuring the quality and readiness of a system prior to its widespread deployment. In this study, testing was applied to the Knowledge Management System of Desa Digital Indonesia (KMS DDI) using three approaches black-box, white-box, and performance testing, following the Software Testing Life Cycle (STLC). Black-box testing was conducted using Katalon Studio through a combination of manual and automated testing with the error guessing technique. White-box testing employed the basis path testing technique using the Jest framework. Performance testing was carried out using Apache JMeter with a load testing scenario ranging from 1 to 500 virtual users. The black-box testing, comprising 312 test scenarios, identified 51 errors with an 84% success rate, most of which were functional errors. The white-box testing identified five errors with a 91% success rate. In the performance testing, most endpoints achieved Good to Excellent Apdex levels, however, several issues were still observed, including errors with the message "connection reset" and slow response times on certain endpoints as well as during initial requests. All identified errors were subsequently analyzed to provide improvement recommendations aimed at enhancing the overall quality of the system.