Metrics
GET /control/v1/monitoring/metrics with Authorization: Bearer <metrics key>, in Prometheus text format. Create the key in Health & alerts → Metrics for Grafana or Datadog. It reads these numbers and nothing else; there’s one per organisation. Answers are cached for 15 seconds.
scrape_configs:
- job_name: onshore
scheme: https
metrics_path: /control/v1/monitoring/metrics
authorization: { type: Bearer, credentials_file: /etc/prometheus/onshore-metrics-key }
static_configs: [{ targets: ["vault.yourbank.example"] }]| Metric | Meaning |
|---|---|
rv_component_up{component} | 1 if the service answered its last health checks |
rv_audit_events_last_hour{module,action,decision} | Audited operations in the last hour |
rv_vault_latency_p50_ms, rv_vault_latency_p95_ms | Tokenize and detokenize time, last hour |
rv_open_alerts | Alerts nobody has acknowledged |
rv_webhook_deliveries_last_day{status} | Webhook deliveries by outcome, last 24 hours |
rv_webhook_oldest_pending_seconds | Age of the oldest undelivered webhook |
rv_siem_lag_events | Audit events not yet sent to the SIEM |
These are gauges over a rolling window, not counters: use them directly, not through rate().
Metrics can’t show Onshore’s own failure. Alert on a failed scrape, and check https://<host>/healthz and the console’s address from outside.
Last updated on