A Qualitative Case Study of Relational Database Index Tuning Using Machine Learning

Abstract

Database performance is a critical factor in determining the application speed. Database indexing is a well- established technique to reduce the query response time, increasing the application speed. The research follows a qualitative analysis approach and aims to drive index tuning to be a dynamic and automated task using ML. This paper is part of the Automatic Index Tuning series and presents the data collection, analysis, and research findings for the index tuning module. The earlier papers in this series presented a literature review, methodology, and theoretical framework. The current paper explains the qualitative analysis process to standardize the parameters influencing the index tuning decision, paving a new path to make index tuning a dynamic and automated task. In addition, it will throw light on the pros and cons of using Machine Learning (ML) classification for index tuning.