System status

Platform health, live

Real component readings from the platform’s own health checks, requested every 30 seconds while this page is open. Nothing here is edited by hand - a green row means the probe passed when the page last asked, and a red one means it failed.

Checking…

Reading live probes…

30s

Reading live probes…

What this page measures

Each row is read from a live probe when the page asks: the database and cache report through Spring Boot’s health indicators, the event bus answers a real broker describeCluster(), disk space is the API host’s own reading, and the embeddings and WhatsApp services answer their own liveness endpoints (liveness only: they do not test model quality or message delivery). The API row is simply this request being answered. A component marked Not deployed is intentionally off in this environment, not failing; Unknown means the probe could not be read, which is not the same as healthy. This page shows only the current reading: it keeps no history and publishes no uptime percentage. API consumers can read the same feed at GET /api/public/status.