Add documentation for the NGINX VTS Exporter to Monitoring docs

We recently added the NGINX VTS Exporter to Omnibus, which means we can get great metrics like throughput, latency, and errors. (omnibus-gitlab!2795 (merged))

We should make sure we document this in the Prometheus Monitoring section:

• We should add an "exporter" section for NGINX, with how to configure it: https://docs.gitlab.com/ee/administration/monitoring/prometheus/#bundled-software-metrics
• We should also update the samples for how to configure an external Prometheus server: https://docs.gitlab.com/ee/administration/monitoring/prometheus/#using-an-external-prometheus-server