Know the moment your site stops answering.

Uptime checks visit your site the way a customer would, from outside your network, and tell you when it stops answering or starts answering wrong.

Four kinds of uptime check

• HTTP: loads a page and checks the response code (any 2xx or 3xx by default, or the codes you choose) and how long it took.
• Keyword: an HTTP check that also looks for a word on the page, like "Add to cart", or makes sure a word like "Error" isn't there. A page can load fine and still be broken.
• Ping: checks that a server is reachable at all. When ICMP is blocked we fall back to connecting on port 443, then 80.
• Port: connects to any TCP port, for mail servers, databases you've exposed on purpose, game servers and the like.

As often as every 30 seconds

How often each site is checked depends on your plan: every 5 minutes on Free and every minute on Paid. On Paid, any domain can have fast checks every 30 seconds, for 10% more on that domain. See pricing.

Confirmed before you're alerted

A single failed check doesn't mean your site is down. The network between us and you may have hiccupped. So when a check fails, Pingerer checks again right away, from other regions on the Paid plan. Only when they agree does it open an incident and alert you. When your site comes back, you get one "recovered" message with how long it was down.

Slow counts too

Set the response time you expect, and how long it must be slow before you hear about it: a few checks in a row, or the average over the last few minutes. We suggest a limit from your site's usual response times. A slow site is marked degraded, not down, so you hear about a site that's crawling before it falls over, without it counting as downtime.

Uptime history and response times

Every site gets an uptime percentage, a response-time chart and a timeline of incidents. Planned maintenance is left out of the numbers when you set up a maintenance window.

Read more: Uptime checks in the help center.