Find out that a cron job never ran before your customers do
Regular monitoring tells you the server is alive. It will not tell you the nightly backup has not started for three weeks. Tickwatch watches your job schedules — and shows the result to your customers on a public page.
- 5 monitors free forever
- no card required
- connect over HTTP
The server is green. There has been no backup for three weeks.
Uptime monitoring checks that a port answers. It has no idea you even have a nightly backup — let alone that it quietly stopped running.
And it says nothing about your jobs.
not when a customer asks you to restore their data.
Connect in three steps
Create a monitor
Describe the schedule — a cron expression, an interval, or on demand. Time zone and grace period are configured separately.
Add one line
Add an HTTP request to your job or scheduler configuration. No separate agent or inbound connections are needed.
Hear only what matters
Tickwatch checks expected signals and run duration. It alerts at your configured failure threshold and sends a separate recovery message.
If it can make an HTTP request, it is already compatible
No SDKs, no agents. One curl at the end of the command and the job is on the map.
Monitoring for the job you need done
Use casesCron job monitoring for missed, failed and stuck jobs
Add start and finish signals to cron. Tickwatch detects missed runs, failures and runtime overruns, then sends an alert.
Backups · PostgreSQL / MySQLMonitor database backups and catch missing backup jobs
Track the start, completion and verification of PostgreSQL and MySQL backups. Get an alert if a backup job fails or never runs.
CI/CD · GitHub ActionsMonitor GitHub Actions scheduled workflows
Detect a scheduled workflow that never started, ran late or failed. Copy a YAML example using a ping URL secret and outcome reporting.
A status page with your monitoring
A public page on tickwatch.dev, with a custom domain from Start. Components follow your monitors; publish incidents and maintenance windows and enable email subscriptions for customers.
This incident was created automatically: the “Warehouse sync” monitor did not report within its expected window. Updates will appear here and go out to subscribers.
Statuses use signals from connected monitors and component rules. Before observations arrive, the page shows no data. Internal providers can be grouped under a service name that makes sense to customers.
Notifications you will not want to mute
Monitoring dies from too many alerts, not too few: after two weeks of noise the team simply stops reading the channel. That is why fighting noise here is part of the product, not a setting.
- Schedule
- 0 3 * * * (Europe/Moscow)
- Exit code
- 1
- Duration
- 4 min 12 s
The tail of stderr right inside the notification is usually enough to understand the cause without opening a laptop.
Prices in roubles, no exchange-rate surprises
A free plan without a card. Paid plans are billed in rubles through Robokassa; the final discounted total is shown before payment.
- 5 monitors
- 7 days of history
- 1 status page
- Run duration charts
- Email and Telegram
- Status page on tickwatch.dev
- One member
- 25 monitors
- 30 days of history
- 1 status page
- Duration charts and slowdown alerts
- Custom domain for the status page
- Slack, Discord, Mattermost, webhooks
- Up to 3 members
- 100 monitors
- 90 days of history
- 3 status pages
- Duration charts and slowdown alerts
- Up to 10 members
- 3 custom status page domains
- Priority support
- 500 monitors
- 365 days of history
- 10 status pages
- Duration charts and slowdown alerts
- Unlimited members
- Up to 10 custom domains
- Team activity log
Personal discounts and promo codes are calculated in your account. Annual plans are charged for the full year.
Hitting a limit blocks creating a new monitor, but never switches off the ones already running. Breaking your monitoring over a late payment is the last thing we would do.
What people usually ask
Uptime monitoring answers “is the service responding?”. We answer “did the job actually run?”. Those are different questions: a server can be perfectly healthy while the nightly backup has not run for weeks because someone deleted a line from crontab.
No. The job makes one HTTP request when it finishes. We need no agent, no daemon and no access to your infrastructure — we never connect to your servers at all.
Use the same timezone as your job scheduler. Tickwatch evaluates cron schedules in that timezone, including clock changes. For runs during the transition hour, check your scheduler’s behavior and set a suitable grace period.
Set a grace period for the expected ping. To monitor duration, send /start and configure a maximum runtime. Exceeding it becomes a separate incident cause.
Ping ingestion runs separately from the dashboard. An independent watchdog checks ping processing, queues and notification delivery. Email and Telegram are also checked regularly. External monitoring detects a missing heartbeat if our watchdog or the entire server stops responding.
Payments in rubles through Robokassa are available. Your final price, including discounts, is shown before payment. Pay once or separately consent to automatic renewal; turning it off preserves paid access. Robokassa registers the NPD receipt through Robochecks for self-employed sellers and emails it to the address supplied at checkout. Payment history and links to registered receipts are available in your account.
After the paid period and grace period, the workspace moves to Free. Existing monitors, email and Telegram keep working. Channels outside Free are disabled; limits for new objects and history retention drop to the Free level. History older than seven days is removed during cleanup. In settings, you can download JSON with configuration and the latest 50,000 events, or CSV with events.
Connect your first monitor
Five monitors free forever, no card. If it does not work out in ten minutes, write to us and we will sort it out together.