I have read the official SonarQube documentation on parallel scan performance (Improving Performance), but I couldn’t find clear information regarding the required computing resources.
Here is my use case:
I want to run 50 scan tasks simultaneously.
Each scan task is relatively small, with only a few dozen to a few hundred lines of code.
In this case:
What server configuration (CPU, memory, database, etc.) would be needed to ensure smooth execution?
Are there any recommended formulas or best practices for estimating the required resources?
I suppose you’re asking for SonarQube Server resource recommendations, rather than build agent recommendations (which was my first impression).
Enterprise Edition($$) offers parallel analysis report processing on the server side. Even so, I would not recommend confusing 50 compute engine (CE) workers.
The way analysis works is that an analysis report is compiled on the CI side, bundled, and submitted to the server. The server queues those reports and processes them in turn. So you can run 50 analyses in parallel. You could run 1000. And they’ll be processed in turn, regardless of the server configuration. And with Enterprise Edition, you can configure multiple CE workers to make chewing through the stack faster.