Short answer: yes—if you’re self-hosting Prefect and using your own Postgres, you can monitor that database with Prometheus via the standard postgres_exporter. If you’re on Prefect Cloud, you don’t have access to our managed DB, so you’d monitor your own infra instead. A few options depending on what you want to monitor: - Monitor the Prefect database (Postgres) - Use postgres_exporter to expose DB metrics, then have Prometheus scrape it. - Example (Docker):

docker run -d --name postgres-exporter \
      -p 9187:9187 \
      -e DATA_SOURCE_NAME="<postgresql://prefect_user:prefect_pass@<db-host>:5432/prefect?sslmode=disable>" \
      <http://quay.io/prometheuscommunity/postgres-exporter|quay.io/prometheuscommunity/postgres-exporter>

- Prometheus scrape config:

scrape_configs:
      - job_name: 'prefect-postgres'
        static_configs:
          - targets: ['<exporter-host>:9187']

- Helm (Kubernetes): prometheus-postgres-exporter - Exporter docs (roles/permissions): postgres_exporter - Monitor Prefect app/runtime metrics - Prefect 3 does not expose a built-in Prometheus /metrics endpoint. - You can: - Instrument your own flows/infra with the Prometheus client to emit custom metrics. - Use OpenTelemetry tracing/telemetry from Prefect for observability (spans/logs), and forward via an OTLP collector to your backend. See: - prefect.telemetry.services - RunTelemetry - Telemetry bootstrap Useful Prefect docs for self-hosting and DB care: - Self-hosted Prefect (v3) - Database maintenance Questions to tailor guidance: - Are you self-hosting Prefect, or using Prefect Cloud? - If self-hosted, how is Postgres deployed (Docker, Kubernetes, managed service)? - Do you want only DB health metrics, or also app/flow-level metrics?

@Brendan Dalpe is there a clear documentation on this?