Database tuning using oracle materialized view for manufacturing industry

• Main Authors: Md Khushairi, Norazah, Emran, Nurul Akmar, Menon, Anil Kumar
• Format: Article
• Language: English
• Published: Machine Intelligence Research (MIR) Labs 2018
• Subjects: Q Science (General)
• Online Access: http://eprints.utem.edu.my/id/eprint/21968/2/IJCISIM_5.pdf; http://eprints.utem.edu.my/id/eprint/21968/; http://www.mirlabs.org/ijcisim/regular_papers_2018/IJCISIM_5.pdf

The need to maintain database performance in Silterra Malaysia is crucial as data produced from complex manufacturing processes must be recorded in timely manner for reporting purposes. Query rewriting using Oracle Materialized View (MV) is one form of corrective action adopted by Silterra in order to tune its database, which is usually affected by problematic SQLs. However, whether MVs are useful in most cases of query rewriting is an open problem. In this paper, the flow of SQL query rewriting process using MVs is presented. Steps to identify problematic SQLs and to rewrite them are given based on DBA’s experience in dealing with database performance issue in this industry. The result of using MVs shown using real fabrication data in Silterra reveals that, even though most MVs perform better than queries without MVs, there are cases that require alternative for MVs.