Optimization of Data Warehouse Architecture to Improve Information System Performance

Data growth is increasing day by day, so the data stored in the data warehouse is increasingly piling up. When data is displayed on the dashboard or information system, performance is slow because the process of loading queries from the data warehouse to the information system will access all the data stored in the data warehouse tables. This causes the speed of loading data on information systems to decrease, so optimization is needed in the data warehouse so that the load process becomes lighter even though data growth is increasing. In this research, a scheduling algorithm will be created in Hadoop whose job is to execute the transform extraction process and load summary data into several tables. Aims to streamline and optimized the Extract, Transform, Load (ETL) process to the data warehouse and reduce the volume of data in one table, then will be indexed according to the primary key in each table so that when data is joined to several tables it can be executed quickly. After testing by querying data with the same goal but different tables, namely tables that are optimized and unoptimized produce a query time of 1.418 seconds, while tables unoptimized have a query time of 2.418 seconds. Well as testing the speed of loading data into the information system by comparing the throughput of systems that are optimized and those that are unoptimized have an average throughput difference of 85%. With these results, it can be concluded that the speed in loading data into the information system has been successfully optimized by looking at this comparison.