What Is Uptime Monitoring?

Quick answer
An uptime monitor is an external service that pings your site on a schedule and alerts you when it fails to respond.
Because checks come from outside your infrastructure, they catch outages your internal monitoring can’t — including DNS and hosting failures.
The limit: a ping proves the site is reachable, not that it works correctly or loads fast.

How does uptime monitoring work?

A monitoring service runs checks from servers spread across the internet on a schedule — every minute is typical. Each check requests your URL and expects a response within a timeout, usually 200 OK. If a check fails, the service retries briefly to rule out a fluke, then alerts you by email, SMS, or push. Because the checks originate outside your network, they catch failures your server-side monitoring would miss: a dead DNS record, a broken CDN, or a down hosting provider.

What is the difference between uptime monitoring and application monitoring?

Uptime monitoring is an external black-box check: can a visitor reach the site? Application monitoring is an internal white-box view: what is the app doing, and what’s going wrong? They answer different questions. An uptime monitor tells you the site is down; application monitoring tells you why. You need both: the uptime monitor to catch the outage, the app monitor to find the root cause.

What makes a good uptime monitoring setup?

Pick a service that checks from multiple locations (so one region’s network problems don’t look like your outage), lets you set the interval (one minute for production, five for side projects), and alerts on the channel you actually check — phone push beats an email inbox nobody reads. Add a status page so users can see the state without emailing you. Free tiers like UptimeRobot’s cover a small project’s needs completely.

Where this bites vibecoders
A vibecoded app can be ‘working’ for months and then silently down for a weekend with nobody noticing, because nothing watched it from the outside. The AI assistant deploys the app and stops; it never suggests external monitoring. An uptime check is a five-minute setup — point a service at your URL, configure an alert to your phone — and it converts ‘nobody knows’ into ‘you know in a minute’.

Where AI coding assistants get this wrong

Calling the app deployed without any external monitoring, so outages are discovered by users.
Pointing the check at a page that returns 200 even when broken (e.g., a cached error page).
Choosing checks from a single location, so regional network issues trigger false alarms.
Alerting to an email address nobody checks, which is the same as not alerting.

Checklist

Add an external uptime check to every public site and API you run.
Check from multiple locations with a 1-minute interval for production.
Alert to a channel you actually see: SMS, push, or chat.
Consider a status page so users can self-serve during incidents.

FAQ

Can I monitor uptime myself with a cron job?

Technically yes, but a cron job on the same server can’t catch a hosting or network outage affecting that server. The value of an uptime service is that checks come from outside your infrastructure. If you self-host the monitor, run it from a different provider.

How often should uptime checks run?

Every minute for services where downtime matters, every 5 minutes for side projects. Faster checks cost more on paid plans and add little for most sites, because a 1-minute check still leaves up to a minute of undetected downtime.