Monitoring¶

OIAF exposes health and metrics endpoints for operational monitoring.

Endpoints¶

Endpoint	Purpose
`GET /healthz`	Liveness (returns `{"status":"ok"}`)
`GET /readyz`	Readiness (returns `{"status":"ready"}`)
`GET /metrics`	Prometheus metrics (enable via `OIAF_METRICS_ENABLED=true`)
`GET /version`	Build/version info

The /metrics endpoint is a placeholder in the MVP and will emit real Prometheus series in a future release.

Expected series:

oiaf_access_evaluations_total{decision} — count by decision (allow/deny/challenge/alert)
oiaf_risk_score — histogram of risk scores
oiaf_challenge_total{method,status} — challenges by method and outcome
oiaf_mfa_verify_total{method,result} — MFA verification outcomes
oiaf_http_requests_total{code,method,path} — HTTP request counts
oiaf_http_request_duration_seconds — latency histogram

Suggested alerts:

High deny rate: rate(oiaf_access_evaluations_total{decision="deny"}[5m]) above baseline — possible attack or misconfiguration.
MFA fatigue: spike in challenge.failed / repeated denials for one identity.
Elevated risk: sustained high oiaf_risk_score quantiles.
Control plane down: /healthz failing or scrape down.
Audit chain broken: oiafctl audit verify returns valid:false — page immediately.
Error rate: 5xx rate above threshold.

Structured JSON logs (log_format: json) include request_id, method, path, status, and duration.
Forward to a central pipeline and correlate via X-Request-ID.

Build dashboards for evaluation volume by decision, risk score distribution, MFA success rate, and API latency.