Blog / friday roundup
Five uptime tools for small sites
Five long-trusted monitoring tools that tell you a small site is down before your customers do, and what's actually worth an alert.
This Friday's list exists because of one sentence we hear too often: "a customer told me the site was down." You should never learn that from a customer. We watch our servers constantly and publish what we see on our status page, but your site is more than our server: your domain, your certificate, your WordPress install, and your scheduled jobs can all fail while the server underneath hums along happily. These five tools close that gap, and none of them require a budget.
UptimeRobot
The workhorse of small-site monitoring, and the one we suggest first. Point it at your homepage, and it checks every few minutes and emails you when the check fails. That covers the biggest gap for most site owners: simply knowing. One habit worth adopting: monitor a page that actually exercises the site (a real page rendered by WordPress), not just the bare domain, so a database failure doesn't hide behind a cached front page.
Pingdom-style checks
Pingdom is the long-standing name in checks that come from multiple locations around the world and time each step of the page load. The multi-location part matters more than people expect. We regularly see "the site is down" reports that turn out to be one office's network or one region's routing, and a monitor with several vantage points settles that argument in seconds instead of an hour.
StatusCake
Another veteran, and a good fit when you want a bit more than a ping. Its content matching (alert unless the page contains a particular phrase) is the feature we point people at. A hacked or half-broken site often still returns a page, just the wrong one; we wrote about what that looks like from our side of the server. A monitor that checks for your footer text catches the wrong-page case that a simple up/down check waves through.
Cron and heartbeat monitors
Tools in the healthchecks mould flip the direction: instead of the monitor visiting you, your scheduled job pings the monitor when it finishes, and you get alerted when the ping doesn't arrive. This is how you find out your nightly backup quietly stopped running three weeks ago. Anything on a schedule that matters (backups, imports, certificate renewals you script yourself) deserves a heartbeat. Silent failures are the expensive ones.
GTmetrix
Not an uptime tool, but it belongs on the list. GTmetrix takes a performance snapshot of a page: load timings, page weight, and what's dragging. Run it monthly and keep the reports. Sites don't usually get slow overnight; they get slow one plugin at a time, and a stack of dated snapshots shows you exactly when the slide started. It's the difference between "the site feels slower lately" and "the site doubled in page weight in May."
What's worth waking up for
A word of caution from years of carrying pagers: the failure mode of monitoring isn't too few alerts, it's too many. After the third false alarm during a summer thunderstorm, people stop reading the emails, and then the real outage sails by. Our rules of thumb:
Alert immediately on the site being down from multiple locations, on certificate expiry inside two weeks, and on a missed backup heartbeat. Everything else (a slow page, a single failed check, one region blip) belongs in a daily or weekly summary you read with coffee. Two or three well-chosen checks you trust beat fifteen you've learned to ignore.
Set aside twenty minutes this weekend: one uptime check on a real page, one heartbeat on your backup job, one snapshot in GTmetrix. That's the whole starter kit. And if an alert ever fires and our status page is green, open a ticket with the alert's details; the timestamps from these tools are exactly what we need to find the cause quickly.