Skip to content

Cronitor

Cronitor can request a web address on a schedule and track start, completion and failure messages from scheduled jobs.

Visit Cronitor

Which monitor answers your first question?

Use a website or API check if you want Cronitor to request a web address and examine its response. Use a job monitor if you want your scheduled task to report when it starts, finishes or fails. A regular heartbeat is a shorter signal that says a process was active.

A healthy response from the report download page does not tell you whether today’s report exists. Keep the job outcome separate and report completion only after the expected file is saved.

How can you tell a missing run from one that failed?

For a job monitor, set when the task should run and the extra time it can take. A reported failure and an absent message after the schedule and grace period are different signals. Route them to someone who can check the schedule, job output or storage destination.

Keep event payloads small and free of private data. Check what a wrapper captures before enabling it. Keep alerts and any public status updates separate so a test incident does not reach customers.

What counts on your plan, and what can move with you?

Cronitor’s Hacker plan lists five monitors and one dashboard user, without SMS alerts. Business lists separate monthly charges per monitor and per dashboard user. Browser checks and some status page features have their own pricing.

Cronitor can export monitor configuration in YAML, a text file of settings. Keep that file, your schedule and important incident records separately. A settings export does not preserve historical events or reconnect alert destinations at a new provider.

What should the monitor say after three different outcomes?

Draw three timelines for a fictional report job. Use no real monitor, report, alert or live endpoint.

  1. Missing
    Draw the expected start and the end of the grace period with no event. Label it a missed run.
  2. Failed
    Draw a start and a failure when the fictional file cannot be saved. Label it a reported failure.
  3. Complete
    Draw the file saved and checked before the completion event. Add the person who should look at the evidence when either failure occurs.