Service health (control panel)

Two different pages answer “is InnoQualis up?”, for two different audiences.

Page	Who it is for	What it checks
status.innoqualis.com	Customers and the Trust Centre	External probes of the hub and the website, plus published incidents
Control panel → Infrastructure → Service health	InnoQualis platform team only	In-process checks inside the backend: database, vector store, AI integration, storage, API endpoints, email service, background workers

The control-panel report names the AI provider and model, whether outbound mail is configured, storage writability and per-dependency response times. That is operator diagnostics, so it requires a platform login (is_platform_user, TOTP enrolled). Until 2026-09-13 the same report was served on the hub at /status with no login; it is not any more.

---

Read the report

1. Sign in to the control panel and open Infrastructure.

2. The Service health card shows the overall verdict and one row per dependency:

   Status	Meaning
   Operational	The dependency answered within its time budget
   Degraded	Answered, but something is off (for example an SMTP relay that refuses the connection, or workers with no recent activity)
   Down	Did not answer within the budget — a dependency that will not answer is down
   Not configured	An optional integration was never set up on this deployment (for example AI with no API key). Never pulls the overall verdict down

3. The card refreshes every 30 seconds; the timestamp is the backend’s clock (UTC).

What to do with a red row

Follow the runbook for the dependency: Backend down, Pod not ready, and the on-call guide. If customers are affected, open an incident on status.innoqualis.com — the control-panel page is not customer-visible.

For scripts and probes

GET /api/health and GET /api/status/health stay public and unauthenticated: they return a bare liveness answer and nothing about configuration. GET /api/status (the full report) returns 401 without a token and 403 for a tenant user.