Monitor your website's uptime

Uptime in a website's sidebar tells you whether the website is up, how it has been, and what went wrong when it wasn't. Nothing to install: we fetch your homepage from our servers every 5 minutes, on every plan.

What you see

• Current status, how long it has been that way, and the last check. Owners and editors can check now, at most once a minute, or pause monitoring.
• The last 24 hours, an hour to a bar, and uptime over 7, 30, 90 and 365 days. These count time we actually checked, so a new website's figures say since when.
• Response time: how long the homepage took to answer, including any redirects, with the average, fastest and slowest.
• If checks stop for over 15 minutes, the page says what was last seen and how long ago, rather than saying the website is still up or down.
• Latest incidents: every outage with its cause, when it started and how long it lasted.
• Domain and SSL: when each expires. The domain's date comes from the daily site audit when the registry shares it, and the certificate is read once a day. Neither sends an alert.

What counts as down

A failed check is repeated each minute, and 3 failures in a row open an incident dated from the first, so a blip doesn't count against your uptime. If our own connection looks like the problem, no incident opens. A website that has never answered has no incidents or alerts yet: the page says it isn't reachable, and monitoring starts with its first answer.

Alerts

When an incident opens, and when it ends, the website's team gets an email and a Telegram message. Anyone who turned emails off in their account, or muted the website on Telegram, isn't sent them. Turn these alerts off for a website from the Uptime page or in Settings → Reports. A website with more than 10 outages in 24 hours stops sending alerts until it settles; the page still records each one.

Good to know

• Read all of this from code through the read API, or from an AI assistant through MCP.
• Every check runs from Mumbai, India. A problem seen only elsewhere in the world won't show, and one affecting only that route will.
• We check the homepage only, and only that it answers, not what it says. If your firewall blocks robots, allow the user agent SuprAnalyticsUptime.
• These requests aren't visits: they don't run your tracking script and don't count as events.
• Hourly check totals are kept 400 days and incidents as long as the website. Reset data removes both. Checks stop while an account has no access to its dashboard.