<!-- NOTICE: This Issue tracker is for the GitLab Helm chart, not the GitLab Rails application. Support: Please do not raise support issues for GitLab.com on this tracker. See https://about.gitlab.com/support/ --> ## Summary As a follow-up to https://gitlab.com/gitlab-org/charts/gitlab/-/merge_requests/1726#note_478533921 we should decide our next steps for service monitoring. Currently we define ports on the Service as described here https://github.com/prometheus-operator/prometheus-operator/blob/master/Documentation/user-guides/getting-started.md . for `service.type=LoadBalancer` this will add the following ports to the internal LB: | port | target | | ------ | ------ | | 8080 | puma | | 8181 | workhorse | | 8083 | puma.monitoring.exporter | | 9229 | workhorse.monitoring.exporter | The downside of this approach is that all of these ports will be available on a loadbalancer if one is defined. We discussed in the linked issue either creating a dedicated service for these monitoring ports, or removing them in [favor of pod monitors](https://gitlab.com/gitlab-org/charts/gitlab/-/merge_requests/1726#note_477864452). One thing to keep in mind is backwards compatibility to anyone that is using servicemonitors with prometheus-operator though it's not clear if this would be a big impact.