System health
Monitoring queues, telephony, and dependencies.
Overview
Intry health spans Core API latency, carrier error rates (Twilio/Telnyx), pg-boss job queue depth, and third-party status (Clerk, Stripe, Knock).
Daily signals
| Signal | Where to look |
|---|---|
| API errors | New Relic / Railway logs / Sentry |
| Twilio failures | Twilio debugger + Core logs |
| Telnyx failures | Telnyx portal + /telnyx/* webhook logs |
| Worker backlog | pg-boss queue metrics (PostgreSQL) |
| Push delivery | Knock dashboard + device registration counts |
Key endpoints
| Check | Endpoint |
|---|---|
| VN pool health | GET /api/admin/virtual-numbers/health |
| Tenant health | GET /api/admin/tenants/:id/health |
| QNT health | GET /api/admin/qnt-health/health-status |
Playbooks
- Carrier partial outage — enable SMS/voice fallback messaging; coordinate with CS.
- Database migration — announce maintenance; pause destructive jobs.
- Auth provider incident — status banner in admin + portal.
- Telnyx cutover —
tools/intry-core-telnyx/CUTOVER.md
Customer communication
Use approved status page or email template. Include: impact, workaround, ETA, next update time.
Last verified: 2026-06-21 (commit 293c4a7)