<![CDATA[Materialized Query Table (MQT) stores data from frequently used queries so that users can get data without having to re-compute. This can improve system performance by reducing the cost of queries. The data in the MQT must be updated periodically it will not become obsolete when some change being made in the parent table. There are 2 (two) kinds of commonly used update mechanisms, namely full refresh and incremental refresh. Full refresh recompiles all data from the parent table. While incremental refresh only process data that being changed by utilizing staging table. The staging table stores the data changes (delta) in the parent table so that incremental refresh can be performed. This study simulates and compares the performance of full refresh with incremental refresh to know the impact of both of the query execution time and resource usage (I/O and CPU). The data used are original data derived from previous research and dummy data generated by system to support this research. Test results show that incremental refresh increases performance by more than 10 times at query execution time and increases more than 50 times on resource usage compared to full refresh.]]>
Copyrights © 2018
