Schedule & grace window
Tell Uptimewolf when the job should run and how much lateness is normal. Runs inside the window stay healthy; the first one that misses opens an incident.
A scheduled job that stops firing doesn't throw an error. It just goes quiet, and you find out when the backup you needed isn't there. Uptimewolf watches your cron jobs, backups and workers with heartbeat checks and alerts you the moment an expected run doesn't show up.
Works on the free tier - no credit card required
Cron job monitoring (also called heartbeat monitoring, or a dead man's switch) turns ordinary uptime monitoring around. Instead of Uptimewolf reaching out to check a service, your job reaches out to Uptimewolf every time it runs. A quick ping at the end of the script is all it takes.
The point is to catch the failure mode that normal checks miss: a job that doesn't run at all. A crashed backup, a cron entry someone commented out, a worker that died on a deploy - none of these return an error to anyone. They simply stop. With a heartbeat in place, a missing ping is the alert: if the expected check-in doesn't arrive inside the grace window you set, Uptimewolf opens an incident and tells your team.
No agent, no library to install. If your job can make an HTTP request, it can ping a heartbeat.
Add a heartbeat monitor, set the schedule the job runs on and a grace window for normal variation. Uptimewolf hands you a unique ping URL.
Drop our one-line curl snippet at the end of your cron job, backup script or worker so it pings that URL each time it finishes cleanly.
If a ping doesn't arrive in time, Uptimewolf opens an incident and notifies your channels - so a silent failure becomes a message you actually see.
Append the curl call to your crontab entry. If the job fails before it reaches the ping, no ping is sent, and the missed check-in is the alert.
# nightly database backup at 02:30, then ping the heartbeat
30 2 * * * /usr/local/bin/backup.sh && curl -fsS https://uptimewolf.com/ping/your-heartbeat-id > /dev/null
The same pattern works from any language - a backup script, a serverless function, a CI step or a queue worker. Send the ping only on success, and a crash, a non-zero exit or a job that never starts all show up as a missed heartbeat.
The controls that tell a late run apart from a dead one.
Tell Uptimewolf when the job should run and how much lateness is normal. Runs inside the window stay healthy; the first one that misses opens an incident.
No SDK, no agent. A single HTTP request at the end of the job is the whole integration, so it works from cron, systemd timers, CI or code.
Every check-in is recorded, so you can see exactly when a job last ran and spot a pattern of late runs before it becomes a missed one.
Cron jobs, database backups, serverless functions, CI pipelines, batch jobs and queue workers - if it runs on a schedule, it can check in.
A website that goes down is loud. A job that stops running is silent - until you need what it was supposed to produce.
The worst time to learn a backup hasn't run in three weeks is while you're restoring it. A heartbeat tells you the night it first fails.
A queue worker or cron entry that didn't come back after a release leaves no error behind. The missing ping is the only signal you'll get.
Jobs usually drift late before they stop entirely. The grace window surfaces the drift, so you fix the cause before it turns into a miss.
Free tier includes two heartbeats - no credit card - cancel anytime.
Seven notification channels - route any heartbeat to any combination of them.
Managing jobs in bulk? Create and control heartbeats through the REST API
Cron job monitoring checks that a scheduled task actually ran. Instead of the monitoring service reaching out to a server, the job itself sends a ping every time it completes. If an expected ping doesn't arrive within the grace window, the service alerts you that the job failed to run or ran late.
A dead man's switch fires an alert when something stops happening rather than when something goes wrong. A heartbeat monitor is a dead man's switch for scheduled jobs: it alerts you when an expected ping doesn't arrive, which is how you catch a cron job or backup that silently stopped running.
Create a heartbeat monitor, set its schedule and a grace window, then add the one-line curl snippet Uptimewolf gives you to the end of your cron job. The job pings Uptimewolf on each run; if a ping is missed, you get alerted over email, Slack, Discord, Telegram, Microsoft Teams, Google Chat or webhook.
You set a grace window when you create the heartbeat. A run that finishes inside the grace window is treated as healthy; one that misses it opens an incident, so normal variation doesn't cause false alarms but a genuinely stalled job does.
Yes. Anything that can make an HTTP request can ping a heartbeat: cron jobs, database backups, serverless functions, CI pipelines, queue workers and batch jobs. Each heartbeat keeps a last-run history so you can see when it last checked in.
Yes. Heartbeat monitors work on every plan, including the free tier (2 heartbeat monitors, free forever). Paid plans raise the monitor count and start with a 5-day free trial, no credit card required.
5-day free trial. No credit card. Cancel anytime.