Heartbeat Monitoring

Status: Planned / not yet implemented.

Outbound heartbeat pings (dead-man's-switch integration with services such as healthchecks.io, Cronitor, or Better Uptime) are not wired up in the current release. The scheduler does not ping any external URL when a task runs, succeeds, or fails, and there are no heartbeat Prometheus metrics.

What exists today

The configuration parser accepts a heartbeat block on a process for forward-compatibility, but no runtime code reads it — setting it has no effect. Only three fields are defined:

processes:
  critical-backup:
    command: ["php", "artisan", "backup:critical"]
    schedule: "0 3 * * *"
    heartbeat:
      enabled: false   # accepted but ignored
      interval: 0      # accepted but ignored
      grace: 0         # accepted but ignored

There is no url, success_url, failure_url, method, headers, retry_count, or retry_delay field — earlier drafts of this page documented those, but they were never implemented.

What to use instead today

Until heartbeats land, use one of these approaches:

Ping from inside the task. Have your task script curl your monitoring service on success/failure. This works with any provider and gives you full control over payloads and exit-code handling:
```
#!/bin/bash
set -e
php artisan backup:run
curl -fsS -m 10 --retry 3 https://hc-ping.com/your-uuid-here
```
Poll the scheduler over the API. Cbox Init records real per-task execution history and statistics, exposed at GET /api/v1/processes/{name}/schedule and GET /api/v1/processes/{name}/schedule/history. An external check can read the last run time, exit code, and success rate. See Scheduled Tasks for the response shape.

Heartbeat Monitoring

Heartbeat Monitoring

What exists today

What to use instead today

See Also