One Grafana instance for both GCP staging and production
As discussed in the call, right now the Google staging and production sites are isolated, and we don't have a way to monitor the Geo Prometheus metrics coming out of the Google staging environments. We should have one Grafana instance that can talk to both.