Rene Škarabot (2014) Performance analysis of in memory databases. EngD thesis.
Abstract
The choice of the appropriate database type presents an important factor in any project development. On account of their data access speed, memory databases have gradually gained ground against the formerly predominant standard disk databases. Since the majority of significant memory databases support the ACID properties, data loss is prevented even in case of power failure; hence their growing popularity. The thesis will examine and measure the extent to which, in performance, memory databases differ from one another, as well as from the extensively widespread, be it indexed or non-indexed, MySQL disk database. As the purposes of memory and disk databases diverge, several test scenarios will be presented in order to further illustrate the differences between the two. The thesis will first present the data storage media, relational databases and indexes, along with the databases examined in the thesis research. Then, the focus will shift to the presentation of the test environment, namely the hardware and software used in the research. Finally, the thesis will provide an overview of test scenarios, followed by the analysis of the obtained measurements, as well as the potential improvements.
Actions (login required)