Sizing Installation

Hi,

I have reviewed the topics on sizing on this forum, but they have not answered my query. I understand that sizing is not an exact science, but some guidelines are required on small-scale (small team) vs Enterprise-scale (large teams) classification.

I am using SonarQube sonarqube-8.4.2.36762.
Assuming I have a VM with 1 vCPUs and 4 GB RAM, 200GB disk.

What would be the classification for a team supporting:-

  1. 20 applications
  2. Each application having 3000 Lines of code (LOC),
  3. 5 applications actively built at any one time
  4. Maintaining 3mths worth of data to manage the quality of a codebase over time.

What would be the maximum sizing (ie No of projects & LOC based on 1 project 3000 LOC) before a “local instance of SonarQube with embedded H2 database and local elastic search storage path” implemented on a VM with 1 vCPUs and 4 GB RAM, 200GB disk develops performance issues.

Any suggestion / recommendation is greatly appreciated.

Thanks
Lakshman

Hi, the first recommendation I can give you is to not use H2 for SonarQube in production. You won’t be able to upgrade, ever. Pick one of the supported DB.

Hi Pierre,

Thank you for the response. Greatly appreciated.

I understand the limitation of using H2. I also understand the scalability and limitation associated with using a local instance of SonarQube with local elastic search storage path.

Our requirement is fairly simple.

  1. Our projects are small in size (3000 loc).
  2. We have 20 projects
  3. We have a 5 member team.
  4. We only require reporting on every build.
  5. We are comfortable to discard build reports during upgrade.

I also understand Colin from SonarSource comment:-
SonarQube is a platform that helps you manage the quality of a codebase over time – it’s not meant to be a “get my answer and delete” kind of service.

I understand, an Enterprise, must follow SonarSource recommended approach.

Based on our simple requirement, are you able to advice, how many projects can be supported concurrently on a local instance of SonarQube on a VM with 1 vCPUs and 4 GB RAM, 200GB disk, before it becomes unusable and grinds to a halt.

For example, given the above hardware configuration, would I have issues with my requirements listed above?

Thanks
Lakshman