// use case: cron monitoring
Cron job monitoring with HTTP heartbeats
Silent cron failures are how billing runs get skipped and backups rot. PingAnywhere gives every scheduled job a ping token: touch it on success, get paged when it goes quiet.
01
One curl line per job
Nothing gets installed on the server. If the job can reach the internet, it can report. Works with cron, systemd timers, GitHub Actions, Celery, and Sidekiq.
*/15 * * * * curl -m 10 https://ping.pinganywhere.com/cPAT_OK/start && \ ./run-billing.sh && curl -m 10 https://ping.pinganywhere.com/cPAT_OK
02
Grace periods that match reality
Jobs with variable runtime get a configurable grace window, so a slow-but-healthy run never pages you at 3 a.m.
03
Missed runs and duration tracking
Get alerted when a job never starts, and see run-duration trends to catch the job that gets 5% slower every week.
Unlimited heartbeats come with the $19 flat plan, with no per-job meters.
Frequently asked questions
How do I monitor a cron job that fails silently?
Give the job a ping token: hit /start before the work and the plain token after it. If the job crashes, is never scheduled, or the machine dies mid-run, the success ping never lands and you are alerted within one interval.
What happens if my job legitimately runs long?
Set a grace period for the check so a slow-but-healthy run does not page you. The missed-run alert fires only when the window passes with no heartbeat.
Which schedulers are supported?
Any that can make an HTTP request: cron, systemd timers, Kubernetes CronJobs, GitHub Actions, Celery beat, Sidekiq, Airflow, and queued jobs in any language.
Do cron heartbeats count as monitors?
Heartbeats are unlimited and included in the flat price, separate from your HTTP check count. A team with 40 jobs pays the same $19 as a team with one.