Tag: sysadmin

An AI coding agent rebuilding the same tag all day fills a disk fast. The mechanism, the truth about volume pruning, and a systemd timer that fixes it.

Creating a RAID array is the easy part. The drama starts six months later when one drive starts dying. Here's the survival guide for everything that happens after mdadm --create.

Drive is dying. Which tool do you reach for? ddrescue clones, TestDisk repairs partitions, PhotoRec carves files. They're a workflow, not rivals.

Containers are just namespaces and cgroups in a trench coat. Build one yourself with unshare and nsenter, no Docker required. Here's what actually happens.

PID 1 zombie reaping in containers, tini, dumb-init, and docker --init compared; when each one fixes your signal handling and stops your 10s shutdown tax.

Your RAID 5 rebuild on a modern multi-TB drive has a 40-50% chance of hitting a URE before it finishes. Here's the 2026 math and what to do about it.

You've been compromised. Now what? A practical incident response playbook for self-hosters who didn't think they'd need one until right now.

Twenty powerful bash one-liners every sysadmin should know, file ops, process hunting, networking, text processing, disk analysis

Compile software on Raspberry Pi or cheap VPS with 512MB, 2GB RAM. Swap, parallel jobs, ccache, and swappiness tuning make it work.

iptables is being phased out. nftables is faster, cleaner, and already the default on modern Linux. Here's how to actually use it without wanting to quit.

Linux ships with conservative kernel defaults. These sysctl settings tune your server for networking, memory, and file I/O, with explanations, not just paste-and-pray values.

Cockpit is the modern systemd-native Linux admin panel. Webmin is the veteran that configures everything. Here's which one should be on your servers, and which shouldn't.