customers.GitLab.com should have Prometheus monitoring

This host keeps running out of disk space because it is not being monitored correctly: https://gitlab.com/gitlab-com/gl-infra/infrastructure/-/issues/6705

Incident: production#2342 (closed)

At a minimum, we should install node-exporter across the customers.GitLab.com fleet.

Once that is done, and assuming the correct type labels are applied, we will automatically start getting utilisation alerts, for high CPU, disk utilisation issues, etc.