Boost Linux Server Performance: Essential Kernel Parameter Tweaks and Resource Limits

1. Optimize Kernel Parameters

Modify /etc/sysctl.conf to add parameters. Use sysctl -a to view defaults and sysctl -p to apply changes.

sysctl -a
sysctl -p

Network‑related parameters such as net.core.somaxconn, net.core.netdev_max_backlog, net.ipv4.tcp_max_syn_backlog, net.ipv4.tcp_fin_timeout, net.ipv4.tcp_tw_reuse, and net.ipv4.tcp_tw_recycle can be increased to improve connection handling.

net.core.somaxconn=65535
net.core.netdev_max_backlog=65535
net.ipv4.tcp_max_syn_backlog=65535
net.ipv4.tcp_fin_timeout=10
net.ipv4.tcp_tw_reuse=1
net.ipv4.tcp_tw_recycle=1

Typical tuning values (selected parameters):

fs.file-max: default 1048576 → tuned 9999999 – total file descriptors.

net.core.rmem_max: default 124928 → tuned 8388608 – max TCP receive buffer.

net.core.wmem_max: default 124928 → tuned 8388608 – max TCP send buffer.

net.ipv4.tcp_keepalive_time: default 7200 → tuned 900 seconds.

net.ipv4.tcp_keepalive_intvl: default 75 → tuned 30 seconds.

net.ipv4.tcp_keepalive_probes: default 9 → tuned 3.

Note: larger values are not always better; consider hardware capacity and the impact on other services.

2. Increase Resource Limits

Edit /etc/security/limits.conf to raise the number of open files.

* soft nofile 65535
* hard nofile 65535

Explanation of fields: * applies to all users, soft is the current limit, hard is the maximum, nofile is the max open files, and 65535 is the new limit. Changes require a reboot.

Case study: Nginx experienced “Too many open files” errors during peak traffic because the default open‑file limit was only 1024. After checking ulimit -a and limits.conf, the limits were raised and Nginx configuration added: worker_rlimit_nofile 65535; Verification screenshots show the new limit applied.

3. Disk Scheduling Strategies

Scheduler can be set via /sys/block/<em>devname</em>/queue/scheduler. Common algorithms:

noop : simple FIFO, best for SSDs.

deadline : guarantees service within a deadline, good for databases.

anticipatory : adds a short wait after reads to merge writes, suited for write‑heavy workloads.

cfq : completely fair queueing.

Relevant kernel parameter directories:

/proc/sys/abi/* – binary compatibility.

/proc/sys/fs/* – file limits and quotas.

/proc/sys/kernel/* – PID limits, hot‑plug, shared memory.

/proc/sys/net/* – network tuning.

/proc/sys/vm/* – cache and buffer management.

Images illustrate the Nginx limit check and verification.