Rasmus Berggren, Dennis Londögård:
Benchmarking and comparison of a relational and a graph database in a CMDB context,
summary, report, June 2020.

Abstract:

The IT company Axis has grown a lot in the past decade and they are now having trouble keeping track of everything efficiently and hypothesize that they are in need of a CMDB, while also wondering what type of database it would be best implemented as. The goal of this thesis was to investigate what requirements Axis has for a CMDB, and use their requirements to analyse if it would be best implemented as a graph database or relational database.

The work was done in two parts. The first part consisted of making a requirement specification for Axis context, this was done with a combination of literature study, interviews as well as examining an old database used for CM purposes. A minimum viable product was created from the requirements to serve as a proof-of-work as well as a basis for the second part. The second part was a benchmarking based on the parameters of performance, writability, as well as readability. A combination of these three parameters was used to answer which type of database would be best for a CMDB.

By conducting the interviews we were able to create a CMDB requirement specification for Axis. The resulting specification suggests that Axis would benefit from an interconnected CMDB between the departments with a need for traceability. The results from the benchmarking showed that relational databases perform better than graph databases, while graph databases were better in regards to writability and readability.

We conclude that Axis would benefit from a CMDB within the company. Our research shows that both of the database types could be used for a CMDB, however we suggest that it should be implemented as a graph database since it would be easier to maintain.