Questions about monitoring

A persistent monitor with schedule, state and configured alerts. Site monitors use active checks; routines send check-ins. We keep the schedule, detect missing runs and send configured alerts. Your agent can stop after setting up the monitor.

Monitor a job or siteUse with an agent

HookPulse does both. A routine sends a check-in to its private HookPulse link. If the check-in does not arrive by the expected time, HookPulse sends an alert. This detects backups, scheduled tasks and automations that stopped running. A site is the other way round: HookPulse calls its address every few minutes and sends an alert when it stops answering and when it is back.

We call the address from our server. The site is up when it answers with a code from 200 to 399, or the codes you choose, and the page has your text, if you set one. Two failed checks in a row, one minute apart, confirm the fall. We also confirm that our own connection works before we alert. One good answer ends the incident and sends the recovery message.

We check monitors every five minutes. If a check-in is late, we send the alert to your configured destinations. The check can add up to five minutes to the delay you allowed. Unresolved alerts repeat once a day by default. You can change that repeat interval. A successful check-in clears the incident.

Yes. Choose a schedule instead of an interval. Set the time zone and the allowed delay. For example, a backup scheduled for 03:00 with a 90-second allowance is late at 03:01:30. The technical reference explains cron expressions and clock changes.

Yes. Connect the failure signal shown in the integration instructions. A start signal records when the run began. A failure signal records the problem. Only a successful check-in resets the expected completion time. Failure alerts are sent on the next five-minute check.

Set a maximum run time on the monitor. Connect the routine's start signal so HookPulse can measure that time. If the run stays open beyond the limit, the next check sends an alert. You can set a limit from one minute to 24 hours.

Once a day by default. You can set a repeat interval from one hour to 30 days for each monitor. The repeat interval cannot be shorter than the monitor's check-in interval. The panel shows the interval that applies.

The monitor shows recent run times, their middle value and the longest run. It measures runs that send both a start signal and a completion signal. Compare those values over time to see whether the routine is taking longer.

Give each run its own identifier. Send the same identifier with its start and completion signals. HookPulse can then measure each run separately. The integration instructions show how to add the identifier.

Yes. Prometheus can read check-in age, overdue status and missed-check counts from your status link. Reading those measurements has no usage charge. The link grants access to the data. Keep it private and replace it if it leaks.

The alert identifies the routine and its last check-in. You can receive it by e-mail or through a connected app such as Slack or Discord. Integration details and the message format are available in /developers.

Yes. After an alert, the next successful check-in sends a recovery message to the same destinations. It includes the start, end and duration of the incident. The technical reference has the message format for integrations.

A request to the monitor's basic check-in link records success and resets its expected time. Start and failure signals have separate links and do not reset that time. Check-ins are free. Open the integration instructions for your tool's example.

Ten monitors are free, with their alerts. Each extra monitor costs US$ 0.10 once, with its alerts, and moving a monitor to an interval under 90 seconds costs US$ 0.05. An internet network can create up to 500 monitors. During the 7-day free trial and within a plan, extra monitors and short intervals cost nothing. During the 7-day free trial, all monitors send alerts. After the trial, alerts come from the ten oldest monitors and from each monitor that you paid for. Alerts from the other monitors need a plan. When the trial or a plan ends, we send one e-mail with the plan that covers all your monitors. The alerts on hold go out right after the purchase. Without an account, alerts from the other monitors wait until you sign in. The monitor page shows when its alerts are on hold. E-mail alerts to an address you choose are free on one monitor. Each additional monitor with such alerts costs US$ 0.10 once, even when you reuse the address. Alerts have no charge per message. See Pricing for payment options.

Yes. Get a setup key and give each routine its own name in the check-in link. Its first check-in creates the monitor. Later check-ins with that name use the same monitor. Service limits still apply. Open Integrations and agents for the link format and key replacement instructions.

Paste your cron schedule into the import tool. Review the proposed monitors and the lines it could not import. Confirm the proposal to create the monitors. The preview does not create anything.

Yes. Add the HookPulse step at the end of the job and let the job request an identity token. The first run creates the monitor with the workflow's schedule and one hour of allowed delay. To receive alerts in an app such as Slack, set its webhook address in the step. To receive e-mail alerts and manage the monitor in your account, add your setup key as a repository secret. Without changing any workflow, install the HookPulse GitHub App: a missed or failed run then opens an issue in the repository.

Pause the monitor and set when the pause ends, up to 30 days ahead. It sends no alerts during that period. You can resume it earlier. The last check-in stays recorded. Keep the routine's check-in connection in place.

You can start without an account. This browser keeps access to your monitors. Sign in to keep them with your account and use another device. E-mail alerts on additional monitors remain paid.

Yes. Choose one of four weekly packages at /partners. Select a period from the catalogue or subscribe with a weekly card charge. Check the total and upload your artwork. Payment precedes approval. Cancel renewal from your account. Partnerships and custom campaigns use the separate form.