20 Proven Linux Server Performance Tweaks Every Sysadmin Should Know

Linux servers are globally renowned and differ from Windows mainly by defaulting to a command‑line interface without a GUI, focusing on handling high‑load, non‑interactive processes efficiently.

High‑availability Linux clusters enable multiple machines to work together, providing fault tolerance and allowing maintenance without service interruption.

Linux servers serve many roles such as web, CMS, file, VoIP, mail, DNS, database, and cloud‑infrastructure nodes.

This article discusses practical tips to improve Linux server performance.

01 Adjust Linux kernel elevator algorithm for disk I/O

Choosing the right filesystem and tuning kernel mount options, especially the elevator algorithm, helps balance low latency and efficient disk read/write request ordering.

02 Disable unnecessary daemons

Many daemons run by default consume memory and CPU without providing value, and each extra service widens the attack surface. Disabling them speeds up boot, frees resources, and enhances security.

Daemons that should be disabled:

Apmd – Advanced Power Management daemon

Nfslock – NFS file locking daemon

Isdn – ISDN modem support

Autofs – Automatic filesystem mounting

Sendmail – Mail Transfer Agent

Xfs – X Window font server

03 Turn off GUI

Linux servers typically do not need a GUI; set the init level to 3 (command‑line) instead of 5 (graphical) and start X only when required.

04 Clean unnecessary modules or features

Review server packages (e.g., Apache modules) and disable unused features such as FrontPage support to free memory and improve speed.

05 Disable control panels

Popular panels like cPanel, Plesk, Webmin, and phpMyAdmin can consume ~120 MB of RAM; disable them unless truly needed, reducing memory usage by 30‑40%.

06 Improve Linux Exim server performance

Use a DNS caching daemon (e.g., djbdns) to lower DNS lookup bandwidth and CPU time, enhancing mail server throughput.

07 Use AES256 to strengthen GPG file encryption

Encrypt backups and sensitive data with GPG using the AES256 algorithm, offering strong 256‑bit security endorsed by the NSA.

08 Remote backup service security

Choose providers that transfer data via SCP or rsync over SSH, preventing direct access to backup servers and protecting against data loss or unauthorized deletion.

09 Update default kernel parameter settings

Adjust kernel parameters (e.g., shmmax, msgmni, file‑max, sem) to support high‑concurrency database workloads.

kernel.shmmax=268435456 (32‑bit) kernel.shmmax=1073741824 (64‑bit) kernel.msgmni=1024 fs.file-max=8192 kernel.sem="250 32000 32 1024"

10 Optimize TCP

Increase TCP window sizes for high‑latency, high‑bandwidth WAN links to improve throughput.

11 Choose the right filesystem

Use ext4 instead of ext3:

Ext4 expands storage limits.

Provides journaling for data integrity.

No disk check needed after unclean shutdowns.

Faster writes due to optimized disk‑head movements.

12 Use noatime mount option

Adding noatime to fstab improves performance, especially on external storage.

13 Adjust Linux file descriptor limits

Default per‑process limit is 1024 descriptors; increase it with ulimit for high‑concurrency servers.

14 Properly configure MySQL

Allocate appropriate memory to MySQL caches; increase cache size when load rises, or reduce it if memory is constrained.

15 Properly configure Apache

Tune StartServers and MinSpareServers to free 30‑40% memory.

16 Analyze Linux server performance

Identify bottlenecks by monitoring CPU, memory, disk I/O, and network usage; use tools to pinpoint slow applications or hardware limits.

17 Learn 5 Linux performance commands

01 top

Displays real‑time process, load, and memory statistics, updating every few seconds.

02 vmstat

Shows snapshots of CPU, I/O, process, and memory usage.

$ vmstat 10

03 iostat

Reports CPU utilization, device I/O, and NFS statistics; options -c, -d, -h isolate each report.

04 free

Shows RAM and swap usage; -s repeats output at a set interval.

$ free -s 5

05 sar

Collects and records performance data over longer periods.

18 Move log files to memory

Store logs in RAM (e.g., using ramlog) and copy them to disk on shutdown to reduce disk wear and improve battery life on laptops.

19 Pack before writing

Allocate a fixed RAM area for logs, writing to disk only when necessary; on SSDs, reserve 50‑80 MB for ramlog to extend drive lifespan.

20 General tuning tips

Prefer static over dynamic content, use FastCGI or mod_perl instead of CGI, and select the most efficient APIs for dynamic applications.