Health endpoint
GET /api/v1/health is unauthenticated and returns the application name, API version, supported locales and enabled panel drivers. It is the right target for an uptime check and for verifying a deployment.
curl -s https://api.example.com/api/v1/healthMetrics
Metrics are exposed for scraping and guarded by MONITORING_METRICS_TOKEN. MONITORING_METRICS_WINDOW_MINUTES sets the aggregation window and MONITORING_METRICS_RETENTION_SECONDS how long samples are kept.
Alerts
| Setting | Fires when |
|---|---|
| MONITORING_ALERT_QUEUE_BACKLOG_THRESHOLD | Queued jobs exceed the threshold — usually a stopped worker. |
| MONITORING_ALERT_JOB_FAILURE_RATE_THRESHOLD | Job failure rate crosses the threshold — often a panel or gateway outage. |
Alerts are written to MONITORING_ALERT_LOG_CHANNEL and POSTed to MONITORING_ALERT_WEBHOOK_URL if set.
TipQueue backlog is the single most useful alarm on this platform. Provisioning, billing and email all run through the queue, so a stopped worker is silent from the outside until a customer complains.