# How do I configure the uptime check timeout?

By default, Oh Dear considers an uptime check to have failed if the response doesn't come back within a few seconds. You can raise or lower that threshold per monitor.

## Where to change the uptime check timeout

1. Open the monitor.
2. Go to **Settings** > **Uptime check**.
3. Find the **Timeout** field.
4. Enter a value (in seconds).
5. Save.

## What counts towards the timeout

The timeout covers the **entire** request: DNS lookup, TCP connect, TLS handshake, sending the request, waiting for the server's first byte, and reading the response body. If the whole round-trip doesn't complete within the limit, we mark the check as failed.

## Uptime check timeout recommendations

- **Default (5 seconds)** works for typical HTTP endpoints with healthy origins. If your site takes longer than this to respond, users are already struggling.
- **Raise it for heavy endpoints** like admin dashboards, search pages, or anything doing significant server work. 10-15 seconds is reasonable.
- **Don't raise it beyond 30 seconds** for uptime checks. If your site is that slow, something's wrong, and a 60-second wait isn't going to help the visitor experience.
- **If you're monitoring an API that streams a response**, consider a shorter timeout combined with a response-body assertion on an early token (so we fail fast if the stream stalls).

## Timeout tuning is not a substitute for performance

If you keep raising the timeout to avoid false-positive alerts, the real problem is response-time drift. Use the [performance check](/docs/features/uptime-and-performance-monitoring) to get notified when response times trend up, and fix the underlying slowness rather than hiding it behind a higher timeout.