How to Find the Right Error in the systemd Journal Using journalctl
Filter messages by service, boot, and time to analyze a specific failure without dumping the entire journal and losing critical lines.
Аккаунт в General iT
Войдите или создайте аккаунт для управления услугами в личном кабинете.
Страны не найдены. Попробуйте другое название или код.

One point for every published article.

Filter messages by service, boot, and time to analyze a specific failure without dumping the entire journal and losing critical lines.


How to read the systemd service state, distinguish process startup from application readiness, and gather facts before changing the server state.


We explain Linux load average: why it cannot be read as CPU percentages and how to correlate it with the number of processors and resource wait times.


We analyze the situation where Linux reports no space despite free disk capacity: how to check inodes and find directories with a high number of files.


How to identify the filesystem running out of space and compare its metrics with directory sizes without blindly deleting data.


Check the user, schedule, and Cron environment without manually re-running a potentially dangerous task or modifying crontab.


How to link a site path to a mount point and block device without confusing disk size, partition, and file system.


First look at the query plan: selected indexes, row estimates, and access order. We explain why EXPLAIN ANALYZE requires a separate solution.


Compare current connections, active threads, and the server limit to identify the cause of failures before increasing the connection limit.


Examining CLI and PHP-FPM: How to Check the Command-Line Environment and Understand Why Its Settings Do Not Guarantee Web Request Configuration


How to determine the version and build parameters, verify the configuration, and distinguish between a successful file check and the actual application of settings.


How to verify tar.gz before extraction: gzip and tar commands, exit codes, and five real-world examples. Why a file list and even exit code 0 do not guarantee a backup is usable.

Public pages use basic statistics. Our recording and Yandex Webvisor require your permission. You can keep basic statistics or disable all analytics.
Operator: General IT Rus LLC, INN 2310208877. Statistics and recording help improve the website. Basic statistics: pageviews, a random daily ID, a public URL and time. The ID lasts until Moscow midnight. Original records: up to 48 hours; totals: up to 7 years. Without Webvisor permission, Yandex receives server statistics without visitor IP, User-Agent or cookies. If selected, its browser code collects technical data, cookies, actions and masked page snapshots; recordings last 15 days. Inputs, chat and account areas are hidden. Interaction recording requires permission: clicks, scrolling, technical information and filtered public-page snapshots. Events: up to 7 years; IP and snapshots: up to 90 days. Permission is remembered for up to 365 days; full opt-out remains until changed. Basic statistics only disables recording; Disable all analytics stops new analytics. Previously sent data is not automatically deleted. Withdrawal and requests: support@git.ru. Full terms are linked below.
Your browser requests no tracking: analytics is off.