🩺 API Health Status

Each tile is one of Uno's API backends. A service's health is judged by server errors and availability — everyday client mistakes don't count against it.

• ● Healthy working normally.
• ▲ Degraded some server errors or reduced availability.
• ◆ Degraded (latency) still error-free, but at least one operation's response time regressed sharply versus the previous window. Detected at 7+ day windows.
• ■ Down failing with server errors, unavailable, or getting no traffic at all.
• 🔒 No access the dashboard can't read this service's data yet (a permissions/config gap) — not an outage.

What the numbers mean

• 5xx rate — share of requests the server itself failed (HTTP 500s). This is the true health signal.
• P95 latency — 95% of requests finished faster than this; a quick read on speed.
• requests — how much traffic the service handled in the selected window.
• 4xx (client) — client-side errors like “not found” (404) or “unauthorized” (401). Shown for context; does not make a service unhealthy.