Five Key Nginx Configuration Tweaks to Boost High‑Concurrency Performance

Nginx is one of the most popular web servers worldwide, known for its event‑driven architecture that uses a master process to manage multiple worker processes, each handling many active connections via a non‑blocking event loop.

Nginx high‑concurrency optimization

The following five core directives are critical for tuning Nginx in high‑traffic scenarios:

worker_processes

The worker_processes directive sets the number of worker processes. It can be configured manually or automatically using the auto keyword, which detects the number of CPU cores.

worker_processes auto;

events {
    worker_connections 1024;
}

For a server with four CPU cores, you might set:

worker_processes 4;

events {
    worker_connections 1024;
}

Example for a medium‑size server (8 CPU cores, 16 GB RAM):

worker_processes 8;

events {
    worker_connections 16384;
    multi_accept on;
    use epoll;
}

worker_connections

The worker_connections directive defines the maximum number of simultaneous connections each worker can handle. The default is 1024, but it should be increased based on system resources.

events {
    worker_connections 65535;
}

Maximum theoretical connections = worker_processes * worker_connections . Adjust worker_connections according to CPU, memory, and network bandwidth.

Recommended ranges:

Small servers: 4096 ~ 8192

Medium servers: 16384 ~ 32768

Large servers: 65536 or higher

keepalive_timeout

This directive controls the HTTP persistent‑connection timeout. A typical value is 60 seconds; in very high‑concurrency environments a shorter timeout (e.g., 30 seconds) reduces idle connection overhead.

keepalive_timeout 60;

keepalive_timeout 30;

gzip compression

Enabling gzip reduces the amount of data transferred, improving bandwidth utilization and request handling capacity.

gzip on;

gzip_min_length 1k;

gzip_comp_level 2;

gzip_types text/plain application/javascript text/css application/xml;

File caching

Cache open file descriptors to avoid frequent file‑open operations, which speeds up static resource delivery.

open_file_cache max=10000 inactive=20s;
open_file_cache_valid 30s;
open_file_cache_min_uses 1;
open_file_cache_errors on;

Static file caching example:

location ~* \.(jpg|jpeg|png|gif|ico|css|js)$ {
    expires 365d;
    add_header Cache-Control public;
}

Additionally, increase the file‑descriptor limit for each worker using worker_rlimit_nofile and the system ulimit command.

worker_rlimit_nofile 65535;

By applying these configurations—optimizing process count, connection limits, keep‑alive settings, compression, and caching—Nginx can handle significantly higher concurrent loads while maintaining low CPU and memory usage.

Finally, the author offers a free collection of advanced architecture materials and a comprehensive Java interview guide, encouraging readers to add their WeChat for access.