Observability API
Health, metrics, and alerts endpoints for continuous operations.
reference • updated 2026-03-15
Goal
Provide operational visibility across platform health, fiscal throughput, and alert conditions for fast incident response.
Primary endpoints
| Endpoint | Usage | Recommended cadence |
|---|---|---|
GET /v1/health | global service availability | continuous checks every 30s-60s |
GET /v1/metrics/fiscal | fiscal throughput and error metrics | hourly reviews and dashboards |
POST /v1/alerts | create/trigger operational alerts | when thresholds are crossed |
Minimum metrics taxonomy
- Availability: success rate for critical endpoints.
- Latency: p95/p99 for checkout, webhook processing, and fiscal issuance.
- Backlog: pending events/documents per organization.
- Quality: retry ratio, error classes, and mean time to recovery.
Health read example
curl -X GET https://api.beacon.pt/v1/health \
-H "Authorization: Bearer <access_token>"
{
"ok": true,
"timestamp": "2026-03-15T10:15:00Z",
"services": {
"billing": "up",
"webhooks": "up",
"fiscal": "up"
}
}
Alert policy
- Define thresholds per environment (
test,live) and severity (P1/P2/P3). - Route alerts to explicit operational owners.
- Link each alert to a runbook and closure criteria.
Operational best practices
- Do not mute alerts without incident context.
- Review noise vs. coverage weekly.
- Correlate alert spikes with deploy windows and configuration changes.