PM2 Zero‑Instance Scaling: Auto‑Stop Timeout Configuration

Goal

Reduce hosting costs by allowing a Node.js application managed by PM2 to automatically stop all instances when traffic drops to zero.

Current Constraint

PM2 lacks a built‑in mechanism to trigger a stop after a configurable period of inactivity. Users must rely on external scripts that monitor request rates and invoke pm2 stop or pm2 delete.

Uncertainty

It is unclear whether a future PM2 version will expose a native auto‑stop option, and how permission boundaries affect the ability of monitoring scripts to control PM2 processes on shared hosting environments.

Questions

• Can PM2 be configured to automatically stop all instances after a defined idle timeout without external tooling?
• What permission model is required for a monitoring daemon to safely invoke pm2 stop/start commands on behalf of the application?
• How do the behaviors of pm2 stop versus pm2 delete impact restart latency and environment variable re‑initialization when scaling to zero?