Monitoring

The FERRET module currently provides one monitoring dashboard.

Mongo Overview

Mongo Overview: Mongo/FerretDB cluster overview

This dashboard provides basic monitoring metrics for FerretDB, including:

• Instance status: Running state of FerretDB instances
• Client connections: Client connection count and request statistics
• Resource usage: CPU, memory, goroutine count, etc.
• PostgreSQL connection pool: Backend PostgreSQL connection pool status

Since FerretDB uses PostgreSQL as its underlying storage engine, for more monitoring metrics please refer to PostgreSQL Monitoring.

Metrics

FerretDB exposes Prometheus-format metrics through its built-in exporter on the mongo_exporter_port (default 9216) port.

Key metric categories include:

For the complete list of metrics, see Metrics.

Alerting Rules

The FerretDB module currently uses basic instance liveness alerts:

- alert: FerretDBDown
  expr: ferretdb_up == 0
  for: 1m
  labels:
    severity: critical
  annotations:
    summary: "FerretDB instance {{ $labels.ins }} is down"
    description: "FerretDB instance {{ $labels.ins }} on {{ $labels.ip }} has been down for more than 1 minute."

Since FerretDB is a stateless proxy layer, the primary monitoring and alerting should focus on the underlying PostgreSQL cluster.