A COMPARISON OF SEARCHING DATA WITH, AND WITHOUT ELASTICSEARCH IN A SQL DATABASE

Detta är en Kandidat-uppsats från Mälardalens universitet/Akademin för innovation, design och teknik

Författare: Mehdi Bel Fdhila; [2023]

Nyckelord: ;

Sammanfattning: Digitalization and managing big data are today becoming vital across all sectors of our society. Databases are highly valuable for companies and organizations to ensure that information resources are well managed. As of today, there are diverse database management systems (DBMS) to choose from.Databases are often used in combination with web applications to list or display the data stored. To establish such visualization of data, a connection is required between the database and the web application. The choice of connections is an important aspect for large scale systems. This study compared two solutions for retrieving data from a database. The study was conducted at Karolinska Institutet Biobank in Stockholm where data is currently fetched from an Oracle Database in combination with the Full-Text Search Engine Elasticsearch. The current solution was compared to a new solution where data was fetched directly from the database, eliminating the use of Elasticsearch completely. The objective of this thesis was to perform an experiment to compare performance in the connectivity solutions and see what connection was the most suitable for large-scale system in terms of data authenticity and security. The study involved implementing a back-end spring boot application to retrieve the data directly using Java Database Connectivity Driver. The back-end API was then connected to the existing front-end angular application to compare with current solution. Previous research shows higher performance when performing searches with Elasticsearch, both alone and in combination with a database. The results of this study similarly concluded that higher performancewas found in data retrieval from Elasticsearch. The study also focused on data authenticity and security of the solutions. It was found that reducing the number of components involved in a system by removing Elasticsearch could increase data authenticity and both solutions could be secured by implementing methods to sanitize user input.

  HÄR KAN DU HÄMTA UPPSATSEN I FULLTEXT. (följ länken till nästa sida)