FAQ

How is this different from uptime monitoring?

Uptime monitoring checks a service from the outside — it pings your website or API and tells you if it stopped responding. CronWarden watches scheduled work: jobs that are supposed to run on a timer and report in when they do. And it goes past "did it run" to "did it do its job", by checking the value a run reports and how long it took. Different question, different tool.

Do I need to install anything?

No. A monitor is one line added to your job — a curl to its check-in URL at the end of the run. Anything that can make an HTTP request can check in.

Does a successful check-in mean the job actually worked?

Only that it ran to completion. A job can finish "successfully" and still have done nothing useful. To catch that, have the run report a value and set a rule on it — then a run that succeeds with the wrong number, or no number, is treated as a failure.

What counts as one monitor?

One scheduled job. If a job runs every night, that's one monitor, no matter how many times it runs. Your plan's monitor limit is the number of distinct jobs you're watching.

What happens if I hit my monitor limit?

Creating another monitor — in the dashboard or through the API — is rejected until you remove one or move to a larger plan. Existing monitors keep working.

Is the check-in URL secret?

Yes. The ping key in the URL is long and unguessable, but anyone who has the URL can check in on the monitor's behalf, so treat it like a credential. Status badges use a separate, read-only key, so publishing a badge never exposes anything that can change a monitor.

My job runs somewhere with no outbound internet. Can I still use it?

The check-in is an outbound HTTP request, so the job (or something near it) needs to be able to reach CronWarden. A job on a fully isolated network can't check in directly.

Can I manage monitors without the dashboard?

Yes — the monitors API creates and updates monitors from a script or CI, and you can import an existing crontab in bulk. Your monitoring can live in version control next to your jobs.

Why did I get a "too fast" or "drift" warning?

Those come from duration and value rules you set on the monitor — a minimum duration, or a drift threshold against recent runs. CronWarden only alerts on the thresholds you configure; it doesn't invent them. If a rule is firing on normal variation, loosen it. See Run duration.

How long is my history kept?

Two clocks run side by side. Raw check-ins — every run's timestamp, duration, and reported value — are kept for your plan's history window (7 days on Free up to 18 months on Business). Daily statistics — per-day uptime, run counts, and duration and value ranges — are kept for two years on every plan, so the long-term view of a monitor outlives its raw detail.

Can I filter the monitor list?

Yes — by status, by tag, or by name. Filters combine, so "everything tagged production that isn't up right now" is one click and a keystroke.

Still stuck?

Work through the Getting started guide, or browse the sections in the sidebar — most questions are answered in Check-ins, Monitors, or Alerting.