This study presents perceptions towards non-relational, or just NoSQL, databases in IT- and Business Intelligence companies, where we try to determine the need for NoSQL databases in these areas. Earlier studies indicate that the rise of Big Data will force companies to sooner or later expand their data storages to be able to handle the vast amount of data that will emerge, and to find ways to easily manage it. With this in mind, we wanted to determine why some Business Intelligence-companies still had not implemented non-relational databases in their systems, what this would implicate and how necessary it actually is. Our study included a case study at a Business Intelligence-company that still had not adopted these relatively new kinds of databases. First we used a survey to get an idea of what kind of knowledge concerning NoSQL-databases the employees at the company actually had, and to help us create questions for the interviews that would be the next step in our study. Through the interviews we had with three consultants and two developers we gathered a lot of interesting data concerning the company’s attitude towards implementing alternative databases in the system as a complement to their current SQL-based databases. For example, the low level of knowledge about NoSQL-databases of the employees showed a lot of preconceptions and assumptions about these alternative databases, that by extension led to a very skeptical attitude towards them. We learned however that the employees, especially the consultants, worried about the growing amount of data, which already was becoming laborious, and we estimate that companies that has not adopted a solution for this will eventually reach an inflection point where the amount of data will exceed the company’s ability to handle it. / <p>På grund av rådande pandemi (covid-19) hölls Studentkonferens Examensarbete i informatik 2020 genom videokonferens (Zoom)</p>

There are many types of databases, but the most common are relational and non-relational. These databases have different structures, and that affects their performance. Many studies examine the differences between relational and non-relational databases and compare them regarding performance. However, it lacks a study that collects the results from different sources and makes them available to software professionals, so they can choose a suitable database effortlessly. This thesis examines and analyzes several studies investigating the performance of relational and non-relational databases. The analysis examines the performance of typical database operations, insert, delete, update, and select, on different numbers of records. The results of this study show that the non-relational databases perform better, regardless of the number of records in the database. However, there are some cases where the relational databases perform better. The findings are based on an analysis of seven studies, encompassing databases MSSQL, MySQL, PostgreSQL, Oracle, and MongoDB. / Det finns flera typer av databaser, men de vanligaste är relationella och icke-relationella. Dessa databaser har olika strukturer, vilket påverkar deras prestanda. Många studier undersöker skillnaderna mellan relationella och icke-relationella databaser och jämför deras prestanda. Dock saknas en studie som samlar resultaten från olika källor och gör dem tillgängliga för mjukvaruproffs, för att underlätta valet av en lämplig databas. Denna examensarbete undersöker och analyserar flera studier som utforskar prestandan hos relationella och icke-relationella databaser. Analysen fokuserar på prestandan för vanliga databasoperationer, såsom infogning, borttagning, uppdatering och val, för olika antal poster. Resultaten av denna studie visar att icke-relationella databaser presterar bättre oavsett antalet poster i databasen. Det finns dock vissa fall där relationella databaser fungerar bättre. Resultaten baseras på en analys av sju studier som omfattar MSSQL, MySQL, PostgreSQL, Oracle och MongoDB-databaserna.