Diagnose a Failed systemd Service Without Guessing
Use status, show, journalctl, unit dependencies, and a disposable user service to turn a failed unit into an exact exit reason.
10 min read
Topic
Processes
8 articles
Process creation, execution, signals, and lifecycle behavior.
SIGTERM Before SIGKILL: How Linux Process Shutdown Actually Works
Signals request process actions; a safe lab shows SIGTERM running cleanup with exit 0 while SIGKILL bypasses handlers and yields status 137.
9 min read
Namespaces and cgroups: The Two Boundaries Behind Containers
Namespaces change what processes can see; cgroups account for and limit resources. Inspect both without pretending a container is a tiny VM.
10 min read
Linux Load Average Without Guesswork
Read Linux load average as a queue signal, separate runnable work from uninterruptible waits, and pair the three familiar numbers with CPU count and PSI.
12 min read
What Happens When You Press Enter in a Linux Terminal?
Follow ls -la from a terminal keystroke through shell parsing, command lookup, process replacement, filesystem calls, and pixels on screen.
11 min read
How Linux Pipes Actually Work
Replace the punctuation model of `|` with connected processes, file descriptors, bounded kernel buffers, and backpressure.
12 min read
Read the systemd Journal Without Guessing
Build a reliable journalctl workflow by separating boot selection, unit filters, time windows, priority, and persistent storage.
13 min read
Shell Scripts That Survive Beyond Bash
Keep everyday automation portable by treating the shell as a language boundary, checking tools explicitly, and testing the unhappy path.
10 min read