NoSQL and SQL - we might have a better solution for your Event Store!

One of the most frequently asked questions is how AxonIQ compares to other players such as NoSQL and SQL database vendors in the messaging and event store space.

Users of the open-source Axon Framework often start with available database solutions to store data, messages, and events. In this blog, we share the results of recent tests performed by Digital Frontiers in Germany, one of our consulting partners.

On our weekly calls, we regularly discuss the needs of our customers and partners. We consider these needs from a technical and commercial point of view taking into account information, solutions, and potential new offerings.

A recurring topic is an examination of existing NoSQL and SQL database solutions used with Axon Framework compared to Axon Server Enterprise used with Axon Framework.

It is no surprise that we recommend Axon Server Enterprise or AxonIQ Cloud, right? Yes, in our presentations we compare and show the results based on our own tests.

Digital Frontiers Benchmark about NoSQL and SQL

In order to study these statements and results, Digital Frontiers conducted a series of their own comparison tests without our knowledge.

Employees from Digital Frontiers spend about 20% of their week doing things that they like that might help themselves or the company. Marco Amann from Digital Frontiers in Germany compared Axon Server with NoSQL and SQL databases from both MongoDB and PostgresSQL. 

We appreciate Digital Frontiers and Marco sharing the presentation and results of their study.

A summary of Marco’s findings supporting Axon Server Enterprise:

  • Less processing power needed
  • Less memory used
  • Much higher throughput (events/sec)
  • Significantly less storage

nosql-and-sql

Download Marco's presentation

The AxonIQ Value Calculator

After we learned about Marco’s tests we developed a tool to help Axon Framework users realize the potential commercial impact if we use our own data and data from our end users.

What is the impact when moving away from NoSQL and SQL databases and what can you save on processing power, memory usage, increased bandwidth/throughput, and storage space? The answer is, a lot! 

Compute Cost

  Machine type Number of nodes Monthly fee / machine 5 Year compute cost
Postgres/MongoDB Dedicated
80 Cores
256 GB RAM
2x 3.4 TB NVME
3 € X,X00.00 € XXX,000.00
Axon Server Virtual server
24 Cores
32 GB RAM
1x 256 GB M2 SSD
3 € XX.00 € XX,XXX.00
Axon Server compute savings € XXX,XXX.00

Storage Cost

  Storage per 100M events 5 Year storage Storage fee GB / month 5 Year storage fee
Postgres/MongoDB 105 GB 1,838 GB € 0.XX € XX,XXX.00
Axon Server 28 GB 490 GB € 0.0XX € XXX.00
Axon Server storage savings € XX,XXX.00

Total 5 year savings: € XXX,XXX.XX

Conclusion

So, if you are interested in running your own calculations based on your own data when moving from NoSQL and SQL databases to Axon Server Enterprise as an event store, please don’t hesitate to let us know. We will be happy to facilitate an online, hands-on, exercise with you.

Sign up for a private comparison here. 


Ralf Pelz & Marco Amann