Page 3 - Linux

Several gigabytes of swap usage while free -h simultaneously reports plenty of free memory looks paradoxical, but is usually proactive, intentional kernel behavior rather than a fault. This guide shows how to use Pressure Stall Information, vmstat, and smaps to distinguish harmless from actually problematic swapping, what role swappiness really plays, and when MySQL buffer pool configuration is the actual cause.

Anyone who has set an environment variable in an SSH session only to find it missing inside a cron job knows this problem well. Bash loads a completely different set of startup files depending on context. The load order of profile, bash_profile and bashrc decides which configuration actually takes effect, and this exact detail is where most administrators get stuck.

Anyone who steers systemd services by trial and error wastes valuable time during outages in production. systemctl starts, stops and monitors services reliably, reads status output correctly, and recognizes states like failed or activating instantly and without guesswork. journalctl then delivers the matching, filtered logs for fast troubleshooting in production Linux environments and Magento deployments.

The five fields of a crontab line look simple at first glance, yet lists, ranges, steps, and above all the notorious OR logic between day of month and weekday regularly produce jobs that run at the wrong time or not at all. This guide explains crontab syntax completely, including special characters, environment variables, and timezones.

Standard Unix permissions do not stop a compromised process from reading everything the running user is allowed to read. Mandatory Access Control adds a second, kernel enforced barrier in front of that. This article compares SELinux and AppArmor, explains label based and path based policy with real examples, and shows how to reliably diagnose a MAC caused permission denied.

A failed disk, a destroyed bootloader, or a complete hardware failure usually hits teams without warning. Bare metal recovery is the discipline of rebuilding a server from nothing, including partition table, bootloader, and operating system, with no working base install left to rely on. Teams that plan and rehearse this case in advance need hours instead of days.

The kernel writes every important message, from hardware detection at boot to the OOM killer during live operation, into an internal ring buffer that dmesg makes readable. Anyone who masters kernel log analysis with dmesg finds driver failures, memory problems and hardware defects often minutes before they show up as an application outage, instead of searching for the cause only afterward.

sysctl makes hundreds of kernel parameters changeable at runtime, but a value set only with sysctl -w disappears again on the next reboot. This article shows how changes become permanent through /etc/sysctl.d, which parameters actually matter for a busy web server such as connection backlog, file limits and swappiness, and why blindly copying someone else's tuning guide often does more harm than good.

Servers where sudo is granted blanket ALL=(ALL) ALL access blur the line between a regular user and root entirely and make every privilege grant impossible to control. This article shows how to edit sudoers safely with visudo, how readable aliases enable narrow grants, when NOPASSWD is justified, and how complete logging keeps every privileged command accountable.

Behind every login on Linux, whether via SSH, sudo or a local console, sits a configurable stack of authentication modules. PAM fully decouples applications from the concrete authentication logic and makes it possible to set up login lockouts, password policies and two factor procedures without changing a single line in the application itself.