Data Integrations

Cockpit can pull data from external monitoring sources and display it alongside your Docker metrics. Each integration uses an adapter that knows how to connect, authenticate, and normalize data from a specific platform.

Available Adapters

Adapter	Edition	What it pulls
Prometheus	CE	Metrics (PromQL) + active alerts
Grafana	CE	Alert rules + annotations
HTTP/JSON	CE	Any REST API with field mappings
Datadog	Pro	Monitors, events, CPU metrics
CloudWatch	Pro	Alarms + EC2 metrics
PagerDuty	Pro	Incidents + response analytics

How It Works

1. Create an integration via the API or mobile app
2. Configure the adapter-specific connection settings
3. Test the connection to verify credentials
4. Cockpit polls the source at the configured interval (default 30-60s)
5. Data is normalized into standard types (Metrics, Alerts, Events)
6. View the data in the dashboard or query via /api/integrations/:id/data

Data Types

All adapters normalize their data into these standard types:

Metrics

Numeric values with a name, unit, and labels.

{ "type": "metric", "name": "cpu_usage", "value": 42.5, "unit": "percent", "labels": { "host": "web-01" } }

Alerts

Active alert states with severity.

{ "type": "alert", "name": "High CPU", "severity": "critical", "status": "firing" }

Events

Timestamped occurrences (annotations, resolved incidents, etc.).

{ "type": "event", "title": "Deploy completed", "detail": "v2.1.0 rolled out" }

Creating an Integration

curl -X POST http://localhost:3000/api/integrations \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "adapter": "prometheus",
    "name": "Production Prometheus",
    "config": { "url": "http://prometheus:9090" },
    "poll_interval": 30
  }'

Limits

Edition	Max Integrations
CE	10
Pro	Unlimited
Enterprise	Unlimited