HTTP monitors

An HTTP monitor loads one page on a schedule and checks that it answers correctly. By the end of this page you will have added one. You will also know how to pause, edit and delete it.

Plan: the fastest check interval depends on your plan. Free checks every 5 minutes. Pro Personal goes down to 1 minute, and Pro Business and Pro Agency to 30 seconds.

Add an HTTP monitor#

1. Open the project and click Uptime monitor in the sidebar.
2. Click Add monitor.
3. Leave Type on HTTP.
4. Give the monitor a Name, such as "Front page".
5. Check the URL. It starts with your project's home page. Any address that starts with http:// or https:// works.
6. Pick a Check interval. The list only shows the intervals your plan allows.
7. Optional: type a word or phrase in Keyword (optional). The check then also fails when that text is missing from the page.
8. Pick Alert after: 1, 2, 3 or 5 failed checks.
9. Click Add monitor.

The message "Monitor created. First check runs within seconds." confirms it. The first result shows in the list shortly after.

What the check expects#

Each check loads the URL and follows up to five redirects. The page must answer with status 200 within 10 seconds. Any other final status counts as a failure, such as a 403, a 404 or a 503.

You cannot change the expected status, the timeout or the request method. There are no header or port fields either.

Watch for a keyword#

A keyword catches pages that load but show the wrong content. An error page with status 200 is a common example.

• The keyword must appear somewhere in the page source.
• The match is case-sensitive. "Welcome" does not match "welcome".
• Pick text that stays on the page for good. A keyword from a banner you later remove turns the monitor red.

A missing keyword shows the cause "keyword check failed".

Pick Alert after#

Alert after sets how many confirmed failures in a row come before the alert. With 1, you hear about the first confirmed failure. With 3, the alert waits for the third one in a row.

A higher number filters out short blips, but every step adds one check interval to the wait. The incident still starts at the first failure, so the recorded downtime stays accurate.

Read the monitor list#

Each row shows the monitor's name, its type and its URL. The columns tell you the rest:

• Last 30 days shows one bar per day. Green means at least 99.9 percent of checks passed, and yellow at least 95 percent. Red means fewer, and grey means no checks ran. Hover a bar for that day's figures.
• 24h and 30d show the share of checks that succeeded in each window.
• Response shows the average response time over 30 days, in milliseconds.
• Interval shows how often the monitor checks.

A dot before the name shows the current status. Hover it to see when the last check ran.

Edit a monitor#

1. Click Edit on the monitor's row.
2. Change the Name, URL, Check interval, Keyword (optional) or Alert after.
3. Tick the Alert channels that should hear about this monitor. This list appears once you have added a channel.
4. Click Save changes.

Saving runs a fresh check with the new settings right away. You cannot change a monitor's type. Delete it and add a new one instead.

Pause and resume#

Click Pause to stop the checks, for example while you rebuild a site. Pausing resets the failure count. An incident that was already open stays open.

Click Resume to start again. The next check runs within seconds.

For planned work with a start and end time, a maintenance window is the better fit.

Delete a monitor#

Click Delete on the monitor's row and confirm. The monitor's history and incidents go with it, and this cannot be undone.