Discussion: Initial metrics for self-monitoring project

Problem to solve

We are working to create an "instance administration" project (a.k.a. "self monitoring project") that administrators can use to monitor their GitLab instance. We have much of the background in place (see #26370 (closed)). However, once the project is available, we need to start populating it with useful information.

This ticket is meant to discuss what are the initial charts or metrics we should start with so the project has value to customers.

Further details

The next step that we are currently working on is enabling users to install the project: #32351 (closed)

We plan to put this behind a feature flag until we can have some contents in the project otherwise users would just see an empty project.

Proposal

We have an issue created already for adding the metrics to the project: #29868 (closed)

However, we need to figure out what metrics we could track. The goal would be to identify a small number of initial metrics to include, while also adding the ability to enhance the metrics in the future. We would want teams besides the Monitor:APM team to be able to update these metrics.

For this issue we want to work with groups that are familiar with the metrics used to monitor GitLab instances like GitLab.com and understand what metrics are most valuable that we can then add to the "instance administration" project.

This issue is just for discussion and #29868 (closed) is for implementation.

Links / references

Assignee Loading
Time tracking Loading