Uptime monitoring in detail: what gets checked, how you are alerted, what it costs

This is the long version of the uptime monitoring product page . That page tells you what you need it for. This one tells you how it works.

An outage does not cost you the downtime, it costs you the trust

Two hours of downtime at night are two hundred visitors who saw an error page, an abandoned checkout and an ad campaign that ran into nothing.

You noticed the next morning. From a customer email, not from an alert.

You hear about it first, not last

We check your websites, your shop, your APIs, database ports and hosts at fixed intervals, and we get in touch once several checks agree that something is wrong. On the fastest plan every 10 seconds, on the free plan every 5 minutes.

We do not only check whether a response arrives at all, but also whether the status code is right, whether an expected text appears on the page, and how long the response took.

What exactly gets checked

• HTTP and HTTPS: status codes including your own expectations, response times broken down into DNS, connection, TLS and download, a check for a keyword in the page content, reading values out of JSON responses, following redirects.
• TCP: is a port reachable? For databases, mail servers, SSH and internal services.
• Ping: network availability of a host, as a simple sign-of-life test for routers, switches and internal servers.

Why the confirmation threshold matters more than the interval

Monitoring that keeps crying wolf gets muted after two weeks. From then on it is worthless, at exactly the moment you need it.

That is why a check only counts as “not reachable” once several checks in a row fail or several regions raise the flag at the same time. Short network hiccups and isolated routing problems drop out that way. How many checks or regions have to agree is your own setting, per check.

That also explains why a shorter interval on its own helps very little: if you check every 10 seconds but alert on the first failure, what you mostly get is false alarms. Interval and threshold belong together.

What multi-region really buys you

Measuring from a single point never reveals regional outages. Those are the nasty ones, because nobody tells you about them: the affected visitor simply does not come back.

There are six regions to choose from: EU-Central, EU-North, EU-West, US-East, US-West and APAC. On the free plan we check from one region, on Pro from up to three, from Business from all six. Which regions those are is up to you: the number is capped, not the location. If your customers are in Singapore, you check from there rather than from Frankfurt.

Counting works in slots: a check in three regions occupies three slots and delivers three independent views of the same service in return.

How multi-region works in detail

Where the alert goes

Email is included in every plan, the free one too. From Pro you add Slack, Teams, Discord, Telegram and webhooks, from Business PagerDuty, Opsgenie and SMS.

From Business there are also recurring reminders for as long as an incident stays open, and from Agency escalation levels with an acknowledge button right inside the email.

All alert channels and escalation rules

Three things nobody else thinks about

They run alongside without you having to set up anything extra.

The certificate that expires on a Sunday. We warn 30, 14, 7 and 1 day ahead, so the browser does not end up warning your visitors. To SSL monitoring

The cronjob that died quietly. Your process checks in with us regularly. If the signal stops, we get in touch. Backups, imports, invoice runs. To heartbeat monitoring

The status page that calms your customers. Public, in your design, optionally on your own domain. Saves half the support requests during an incident. To status pages

What uptime monitoring does not do

So you do not expect the wrong thing:

• It tells you that something is broken, not why. We see your site from the outside, the way your customers see it. What exactly is causing it is in your own logs: in the error log of your shop, at your host or in the application itself. Our alert tells you where to look: which check, from which region, which status code and since when.
• It is not a load test. We measure your availability, we do not generate load.
• It does not fix anything. It wakes you up so you can fix it.

Interval and allowance across the plans

Two numbers belong together: the shortest interval you can set, and how many of your checks may run at that fast pace at the same time. Checks that run more often than every 5 minutes are what we call fast checks. Slower is possible on every plan, faster only one step up.

The 30-second pace starts with Agency. Pro and Business share the one-minute pace and differ in the allowance, in workspaces, users and guest seats, in product comparisons and in how far the history goes back. You decide per check which one gets the fast pace: your shop checkout probably, your careers page probably not. All plans side by side on the pricing page .

Frequently asked questions

Do I need technical knowledge? No. Enter an address, pick an interval, done. If you use one of the nine shop and CMS plugins , it even suggests the sensible checks and creates them at the push of a button.

Am I tied in? No. Cancellable monthly. The free plan runs without a time limit and without payment details, including after the 14-day trial of the paid features.

Why not just use a free tool? You can, and for a private site that is often enough. The difference shows up with multiple regions, with alerts that do not constantly misfire, with SSL and cronjobs in the same tool, and with a provider that hosts in the EU and offers a data processing agreement.

Where is my data? In the EU, GDPR compliant. There is a public status page for our own service at status.live24h.eu .

Your first check is running in a few minutes

Start for free → · 15 checks free, no payment details

If you run a shop, price monitoring is worth a look next, and SEO monitoring makes sure Google keeps finding your pages.