I have a Node.js application running inside Docker on a DigitalOcean Ubuntu droplet. The server specs are:

• Ubuntu 22.04
• 2 GB RAM
• 1 vCPU
• Docker Engine latest version
• Nginx reverse proxy
• PM2 inside container

The application works normally for some time, but after receiving moderate traffic (around 40–60 concurrent users), CPU usage suddenly spikes to 100%, memory usage increases rapidly, and sometimes the droplet becomes unresponsive or restarts automatically.

Symptoms

• docker stats shows Node.js container consuming very high CPU
• Nginx starts returning 502 Bad Gateway
• PM2 logs show app restarting repeatedly
• Swap memory also fills quickly
• Occasionally SSH connection to the droplet is lost temporarily

What I Tried

• Increased swap size from 1 GB to 2 GB
• Enabled PM2 cluster mode
• Added Docker memory limits
• Restarted Docker service
• Optimized some database queries
• Enabled gzip in Nginx

Docker Run Command

docker run -d
–name myapp
-p 3000:3000
–memory=“1g”
–cpus=“1.0”
my-node-app

#Questions

1. What could be causing these CPU spikes and random restarts?
2. Is this related to Docker resource limits or Node.js memory leaks?
3. Should I move PM2 outside the container?
4. What is the best way to monitor and debug resource usage on a DigitalOcean droplet?
5. Are there recommended production settings for Docker + Node.js on low-resource droplets?

Any suggestions or troubleshooting steps would be appreciated.