SeekGist status

How we keep SeekGist up

Our goal is 99.9% uptime. In a 30-day month that allows at most 43.2 minutes of downtime (0.1% of 43,200 minutes). This is a design goal, not a measured result.

• Your store never breaks: connectors switch to the store's own search after 300 ms, or at once on any error.
• Health checks: /v1/health shows if the API is alive; /v1/health?deep=true also checks the database, the cache and the search engine.
• Two API servers and a search engine replica, so one server can fail without failed requests.
• Alerts reach the on-call engineer when responses get slow, when errors pass 0.5% or when the AI step times out often.
• This page is a static page that reads the status data in your browser, so it still loads when the API is down.

Measured uptime history

TODO: Not measured yet. It will come from the monitoring system after launch.

Sources: the uptime gate in the build spec and ADR 0012 (connector fallback), ADR 0015 (health checks and alerts).