RECOMMENDATION FOR ENHANCING REPORTING PERFORMANCE OF AN INDONESIAN BANK’S DEPOSIT SYSTEM

• Main Author: Gerald Akbar Giffera, Muhammad
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/85258
• Institution: Institut Teknologi Bandung

Database performance from the query runtime perspective produced by Bank X's system, a bank in Indonesia, is currently considered suboptimal. Based on initial testing, the average query runtime produced is 58 seconds compared to the target of 50 seconds. In this thesis, several alternative solutions are developed, both through relational database optimization and the development of non-relational database solutions to improve the reporting process performance. Optimization of the relational database is done by applying indexes, foreign keys, and database denormalization. The non-relational database solutions consist of data models and adjustments that need to be made to the system for two types of non-relational databases, namely document-oriented and column-oriented. Testing was conducted on four recommended solutions: applying indexes to the relational database (PostgreSQL), denormalizing the relational database schema (PostgreSQL), utilizing document- oriented databases (MongoDB), and utilizing column-oriented databases (BigQuery). Applying indexes in PostgreSQL resulted in a 15.43% reduction in processing time. Schema denormalization in PostgreSQL led to a 58.47% reduction in time. The non-relational database solutions resulted in a 79.1% time reduction in MongoDB and 93.13% in BigQuery.