Skip to main content
Super Admins (level 5) can open System from the Admin sidebar (/admin/system). The page is an operations snapshot of this app’s runtime: whether the API process is up, how large MongoDB is, whether Redis is healthy, and which environment variables are set (values are never shown). System health Caption: Add screenshot showing /admin/system with API, MongoDB, and Redis cards, overall health tag, and Flush caches.

​
What it reports

Overall status:
  • Healthy — API and MongoDB are up; required env vars are set; Redis is up if configured; unique index is present
  • Degraded — Mongo is up but Redis is configured and down, a required env var is missing, or the unique plan index is missing
  • Down — API or MongoDB health check failed
The health API never returns connection strings, Redis passwords, or OAuth secrets. Treat /api/admin/health as Super Admin only.

​
Routes

​
Flush caches

Deletes app cache keys only:
  • qb:published:* (question bank)
  • year:* (school-year settings)
  • public:overview:* (public homepage snapshot)
It does not FLUSHALL, and it does not drop step locks (form:{id}:step:{key}) or save rate-limit keys (rl:*). Writes AuditLog action settings_changed.

​
Retry Redis

Clears the 60-second “Redis unreachable” backoff so the next request tries REDIS_URL again. Use this after Redis comes back without waiting for the timer.

​
Useful later (not built)

These belong in Atlas, Vercel, or a later Super Admin tool — not on the current System page:
  • Atlas backup / snapshot status
  • Vercel deployment and function duration
  • Browse and release a stuck step lock after a crash
  • Duplicate-plan detector if the unique index was missing
  • Slow request / 5xx log beyond optional Sentry (SENTRY_DSN)