[Feature Request] Application health-check (reboot on failure)

What's going wrong?

pm2 is monitoring Node.js processes and restarts them automatically. This works fine but if processes are blocking, e.g. due to a deadlock or other runtime issues, the Node.js process is kept in its blocked state forever. On the OS level it is possible to use a hardware watchdog and for instance a systemd service to verify all processes are responsive. I would like to have such a feature in pm2 to send periodically events from JavaScript or native code to reset a watchdog timer in pm2. If the Node.js process is no longer sending events it shall be restarted.

We are using pm2 on IoT devices and a 100% uptime is crucial. At the moment only crashes are covered but not all other situations.

The watch feature could be a way to implement this feature using a second Node.js app which monitors the first one.

This is not a bug report, this is a feature request.

Start by reviewing PM2’s existing watch feature and how it monitors and restarts Node.js processes. Clarify the heartbeat or watchdog event contract, timeout behavior, and restart conditions; done should include a responsive-process case and a blocked-process case that reliably triggers recovery.