Cronitor is a monitoring service for cron jobs, workers, services, APIs and anything else that can make or receive HTTP requests. Cronitor adds visibility to the critical jobs and backend systems that you rely on.
Learn more about How Cronitor Works
For cron job and heartbeat monitors, you can intergate with Cronitor using any language that can make an HTTP request. The easiest integration
is directly within crontab using
curl, and all modern languages have libraries for making requests. Several great open source API clients
are available, created by our users.
We offer new users a free trial on the plan of their choice. During the trial period all of the plan's features are available.
Yes, one monitor is free! Upgrade at any time to add monitors and unlock premium features.
A monitor is a collection of rules and alert settings that you will create for every job, task, or service that you want to monitor.
Cronitor is a general purpose monitoring tool with a simple HTTP interface.
Yes! Cron job monitors use cron schedule expressions like
0 0 * * Mon-Fri.
If you receive an alert you didn't expect—if you know your job did not run too long for example—the first thing to do is view your recent ping history in your dashboard. Usually the explanation can be found in the history, but please email us if you have any questions. You can read more about viewing ping history, and additional troubleshooting information is available here for timing jobs.
Cronitor ping tracking and monitoring has been engineered to deliver low-latency alerts. As you can see in how cron job monitoring works, we evaluate your monitor's rules continuously and we often send failure alerts within 60 seconds. We know that if your alerts are predictable and precise, you can build your own grace period into your rule and have a clear understanding of when your rules would trigger an alert.
That said, Cronitor does pad rules with a grace period before triggering an alert to reduce false positives. The grace period is small and varies based on the precision of your rule. For example, a not run in 3 hours rule may have a grace period of 1 minute and a functionally identical not run in 10,800 seconds rule will have no grace period at all. If you need more padding to prevent false positives on a job, consider building it into the rule itself.
For example, if you have a "not run in 1 day" rule and you're consistently getting recovery notifications a moment or two after an alert, you may need to relax the rule. Try changing it to "not run in 1445 minutes".
By reviewing your ping history it's often easy to see why you received an alert. If you ping your Cronitor url at the end of your command, the exact time of your ping will depend on how long your command takes to finish. If your "not run in 1 hour" job takes 1 minute to finish, and the next time it takes 3 minutes, your pings will be 62 minutes apart and you're likely to receive an alert. We will continue to evaluate our grace period logic to ensure we're sending relevant alerts in a timely fashion. If you have any thoughts on this, we'd love your input.
Cronitor uses Stripe for our checkout and recurring billing. Your credit card details are never sent to our server. We trust Stripe with our own money and think it's safe for you, too.
Subscribers on our "Monitor Everything" and "Enterprise Cronitor" plans may request to be billed quarterly against a purchase order. All other plans require a credit or debit card for recurring monthly billing. An annual billing option is available to all subscribers upon request. Cronitor provides a 10% discount when paying annually.