Auto-Recovery

PM2 Process Auto-Recovery for Node.js

Detect PM2 process failures and auto-restart your Node.js apps. Works alongside PM2s built-in recovery for extra reliability.

Start Monitoring FreeSee How It Works

Why PM2 Apps Still Fail

  • ✗PM2 itself crashes and takes apps down
  • ✗Memory limits exceeded, restart loops begin
  • ✗Cluster mode workers all fail simultaneously
  • ✗App starts but returns errors (PM2 thinks its healthy)
  • ✗Log files fill disk, causing hangs

How RecoveryPulse Adds Extra Protection

  • ✓HTTP-level health checks (not just process status)
  • ✓Detect when app responds but returns errors
  • ✓Restart PM2 daemon if it crashes
  • ✓Force restart apps stuck in restart loops
  • ✓Alert when PM2s own recovery fails

Sample Recovery Rule for PM2

Force Restart PM2 AppAuto-Recovery
pm2 restart app --force

Force restarts a PM2 app when it responds but returns errors, or gets stuck in a restart loop.

Perfect For

•Production Node.js servers
•Next.js deployments
•API microservices
•Background job processors
•WebSocket servers
•Real-time applications

Frequently Asked Questions

How is this different from uptime alerts?

Most monitoring tools just send you an alert when your site goes down. RecoveryPulse goes further—it SSHs into your server and runs recovery commands automatically, often fixing the problem before you even see the notification.

Is SSH auto-recovery secure?

Yes. Your SSH keys are encrypted at rest using AES-256. We never store passwords. Keys are only decrypted in memory during recovery execution, and all connections use standard SSH protocol with key-based authentication.

Can I customize recovery steps?

Absolutely. You define your own recovery playbooks with ordered actions. Each action can have retry logic, wait times, and conditions. You control exactly what commands run on your server.

What if the automated fix fails?

If recovery fails, we escalate immediately—you get notified via email and SMS with full logs of what was attempted. You can also configure fallback actions or manual-only recovery for critical systems.

How often do you check my site?

We check every 30-60 seconds depending on your plan. When we detect an issue, recovery starts within seconds—not minutes. Most outages are resolved before your users even notice.

Best for VPS and single-server deployments. If you use Kubernetes or ECS, you likely already have orchestrator health checks and replacement.

Start Monitoring Your PM2 Stack Today

Free tier includes 2 sites. No credit card required.

Get Started Free