Cron & heartbeat monitoring
Stop choosing which jobs deserve monitoring.
Nullbeat watches every cron job, backup, and pipeline you run — 500 monitors for $20, flat. When a job goes silent, the right people know in seconds.
20 monitors free · every alert channel · no credit card
Demo: a monitor's heartbeats arrive on schedule, one beat is missed, the monitor flips to down, and a Slack alert is delivered.
One line
Monitoring a job is a one-line change.
nullbeat wrap runs your command and reports the start, the finish, the exit code, the duration, and the output tail. The job itself doesn't change.
0 3 * * * pg_dump acme_prod > /backups/acme.sql0 3 * * * nullbeat wrap db-backup -- pg_dump acme_prod > /backups/acme.sql0 3 * * * pg_dump acme_prod > /backups/acme.sql && curl -fsS --retry 3 https://ping.nullbeat.com/pk_3q6zw8kfyrv9Anything that can make an HTTP request can be monitored. Cron, systemd timers, Kubernetes CronJobs, GitHub Actions, a Raspberry Pi in a closet.
First-class, not a checkbox
Cron monitoring with the details that matter.
Most monitoring products bolted heartbeats on later. Nullbeat is built for scheduled work — so it catches the failures a “did it ping?” checkbox can't.
Catches hung jobs
Start and finish are separate signals. A job that begins and never ends is a failure — even though it “pinged.”
Understands your schedule
Real cron expressions with timezones and grace periods. Nullbeat knows exactly when the next run is due, so it knows the moment one is late.
Records every run
Duration, exit code, and output tail on each execution — when a backup fails at 3 AM, the why is already in the run log.
One alert per incident
Flap detection collapses a bouncing job into a single notification, and recovery closes the loop. No 3 AM pager storms.
When it breaks
Every alert channel, on every plan.
Email, Slack, PagerDuty, Discord, and webhooks — all included on Free. Where your alerts go is not an upsell.
Nullbeat APP 3:32 AM
🔴 db-backup is DOWN (timeout: 24h, grace: 30m)
Status: DOWN
Timeout: 24h, Grace: 30m
Last ping: 2026-07-19 03:02:11 UTC
Org: acme
Recovery notifications close incidents automatically; webhook deliveries are signed and retried.
Monitor everything
Room for every job you actually run.
Most teams monitor the five jobs that scared them most recently. On a per-monitor tax, that's rational. On a flat plan, you monitor all of it.
Every customer's jobs, not just yours
Running scheduled work per tenant? A thousand customers means a thousand syncs that can silently fail. Monitor each one individually — one flat price.
Every client backup, verified nightly
Agencies and hosts: stop promising backups are running and start knowing. One monitor per client site, and the 3 AM failure pages you before the client calls.
Every task in the pipeline, not just the last one
Data teams: when step 14 of 60 hangs, “the report is late” is a bad alert. Monitor each step and know which one died, and when.
Every device, phoning home
Fleets of boxes, sensors, or agents — each one a heartbeat. The one that goes quiet in a closet somewhere becomes a notification, not a mystery three weeks later.
Free covers your crontab. Pro covers your product.
Pricing
Flat tiers. No overages. No per-seat math.
- 20 monitors
- 7 days incident history
- 1 team member
- All alert channels
- API + CLI
- Community support
- 500 monitors
- 90 days incident history
- Unlimited team members
- All alert channels
- API + CLI
- Email support
- 5,000 monitors
- 1 year incident history
- Unlimited team members
- All alert channels
- API + CLI
- Priority support
- 25,000 monitors
- 7 years incident history
- Unlimited team members
- All alert channels
- API + CLI
- SSO + audit log
- Dedicated support
Monitoring never stops because of billing. Paid plans include a soft grace zone past the limit, and hitting a cap only blocks creating monitors — never running ones.
Every new account starts with 14 days of Pro. No credit card.
These aren't promotional prices. We built Nullbeat so that watching another job costs us almost nothing — so we sell monitors the way they cost us: in bulk, flat, without a meter running.
500 monitors runs $80–$1,000/month elsewhere.
FAQ
Questions, answered plainly.
Your crontab is full of jobs nobody's watching.
Twenty monitors free, every channel included, running in the next five minutes.
Start monitoring free