Gremoire

How to Optimize a Linux Machine, VM, or VPS for High Performance ​

In today’s digital infrastructure, whether you’re deploying web apps, APIs, containers, or cloud services, performance optimization is non-negotiable. This article walks you through how to optimize any Linux-based environment — from bare metal to VMs (like KVM/Proxmox) and cloud VPS — to handle multiprocessing, concurrent requests, and high I/O workloads.

1. Understand Your System’s Hardware & Virtualization Layer ​

Check CPU and Memory: ​

bash
lscpu
free -h

⚙️ For VMs: ​

  • Set CPU model to host in your hypervisor (e.g., Proxmox, VirtualBox, VMware) to expose full CPU instruction set.
  • Enable NUMA awareness for large memory systems.
  • Use virtio drivers for better I/O performance.

2. Enable Multi-Core Processing in Apps (e.g., Node.js, Python, PHP) ​

Node.js: ​

Use PM2 for process management:

bash
npm install pm2 -g
pm2 start app.js -i max   # forks processes = number of CPU cores

Python: ​

Use gunicorn with --workers:

bash
gunicorn app:app --workers 4 --threads 2

PHP (FPM): ​

Edit /etc/php/7.x/fpm/pool.d/www.conf:

pm.max_children = 10
pm.start_servers = 4
pm.min_spare_servers = 2
pm.max_spare_servers = 6

3. Tune System Resource Limits (ulimit, sysctl, etc.) ​

Raise File Descriptors ​

bash
# In /etc/security/limits.conf
* soft nofile 65535
* hard nofile 65535

# System-wide
echo "fs.file-max = 2097152" >> /etc/sysctl.conf

Tune TCP Parameters for Concurrency ​

Add to /etc/sysctl.conf:

bash
net.core.somaxconn = 65535
net.ipv4.tcp_tw_reuse = 1
net.ipv4.tcp_max_syn_backlog = 4096
net.ipv4.ip_local_port_range = 1024 65535
net.ipv4.tcp_fin_timeout = 15

Apply changes:

bash
sudo sysctl -p

4. Secure and Harden the System (Without Sacrificing Performance) ​

  • Disable unused services:
    bash
    sudo systemctl disable bluetooth cups avahi-daemon
  • Use a lightweight firewall (like ufw or iptables) and allow only necessary ports.
  • Use fail2ban to protect against brute-force attacks.

5. Optimize Disk & Filesystem I/O ​

For SSD-based VPS: ​

  • Use ext4 with journaling or xfs
  • Mount with options:
    bash
    defaults,noatime,nodiratime

Monitor with: ​

bash
iotop
iostat -xm 2

6. Improve Network Throughput ​

  • Prefer eth0 with virtio or enhanced network drivers
  • Enable TCP BBR congestion control:
    bash
    echo "net.core.default_qdisc=fq" >> /etc/sysctl.conf
echo "net.ipv4.tcp_congestion_control=bbr" >> /etc/sysctl.conf
sudo sysctl -p

Check:

bash
sysctl net.ipv4.tcp_congestion_control

7. Use NGINX or HAProxy for Reverse Proxy & Load Balancing ​

NGINX helps terminate SSL, distribute traffic to app workers, and handle static content:

nginx
upstream app {
    server 127.0.0.1:3000;
    server 127.0.0.1:3001;
}

server {
    listen 80;
    server_name example.com;

    location / {
        proxy_pass http://app;
        proxy_http_version 1.1;
        proxy_set_header Upgrade $http_upgrade;
        proxy_set_header Connection 'upgrade';
        proxy_set_header Host $host;
    }
}

8. Monitor & Profile Your System ​

Use the following tools:

  • htop – Live CPU/RAM usage
  • glances – Full system snapshot
  • iftop / nethogs – Network traffic
  • pm2 monit – Node.js process monitoring

Enable logging and alerts using:

  • netdata
  • grafana + prometheus
  • uptime-kuma (self-hosted status monitor)

Bonus: Automation Script Template ​

Here’s a basic shell script to apply many of the above optimizations:

bash
#!/bin/bash
echo "Optimizing Linux system for high concurrency..."

# Raise file limits
echo '* soft nofile 65535' >> /etc/security/limits.conf
echo '* hard nofile 65535' >> /etc/security/limits.conf
ulimit -n 65535

# Sysctl tuning
cat <<EOF >> /etc/sysctl.conf
fs.file-max = 2097152
net.core.somaxconn = 65535
net.ipv4.tcp_max_syn_backlog = 4096
net.ipv4.tcp_tw_reuse = 1
net.ipv4.ip_local_port_range = 1024 65535
net.core.default_qdisc=fq
net.ipv4.tcp_congestion_control=bbr
EOF

sysctl -p

echo " Optimization complete. Reboot may be required."

Save as optimize-linux.sh, then:

bash
chmod +x optimize-linux.sh
sudo ./optimize-linux.sh

Key Takeaways ​

  • Linux defaults are safe but conservative — you must tune for high loads.
  • Use PM2, NGINX, connection pooling, and sysctl tweaks for scale.
  • Monitor continuously. Measure before and after applying changes.
  • Optimization is not just about performance, but stability under load.