The scripts were attributed to LINUXexpert.org, which is being retired as a
site and is no longer where this work lives. Coffey Labs is the organisation
these projects belong to.
One line per script, fifteen of them, and nothing else. LICENSE is deliberately
untouched: its "Copyright (C) <year> <name of author>" lines are GPL boilerplate
showing you how to write your own notice, and the Free Software Foundation's
own copyright on the licence text is not ours to edit.
Both git contributors are the same person, so there is no third-party copyright
here that could not be restated.
The old README was a flat list of one-line descriptions, several of
which no longer described the script: log_rotate gained flags and no
longer purges by default, restore and process_monitor now confirm before
acting, rsync_magic changed a default, and disk_cleanup grew guards.
Rewritten around the conventions that are now consistent across the
collection -- --dry-run, a confirmation before anything destructive, an
unattended escape hatch, and a refusal rather than a guess when there is
no tty to ask on. That last one is the part worth knowing before putting
any of these in cron.
Groups the scripts by what they are for rather than listing them
alphabetically, and states the things a reader would otherwise have to
discover by reading source: that security_audit as an ordinary user
proves very little, that log_rotate is not a logrotate replacement and
why, that update_system is unattended on every branch, and that
restore's tar options matter because an archive picks its own ownership
and modes.
Also updated three script headers that had gained ASSUME_YES/FORCE
escapes without documenting them, so the headers and the README agree.
Every flag, environment variable and behaviour claimed here was checked
against the scripts rather than written from memory.
set -euo pipefail across all four, but added deliberately rather than
pasted in -- each script needed the places where a non-zero exit is
normal handled first, or strict mode would have made them worse:
- sys_monitor / process_monitor: `ps | head -n 6` is a latent SIGPIPE.
head closes the pipe after six lines, and on a host with enough
processes ps fills the buffer and exits 141, which pipefail turns into
a script abort -- on exactly the busy machine you wanted to inspect.
Confirmed the mechanism (a large producer into head returns 141) and
those pipelines now tolerate it.
- security_audit: find exits non-zero when it cannot read a directory,
which is routine when walking the whole filesystem. Without handling,
set -e aborted the audit part way while still looking complete. Also
notes that a clean report as non-root means little, since find cannot
descend where it may not read.
- network_info: iptables needs root, so the last section aborted the
script for ordinary users. Now reports the failure, and falls back to
nft where iptables is absent.
process_monitor also no longer kills on sight. `pkill -x` by name can
match several processes at once, and as root that is an easy way to take
down more than intended. It now prints what it matched and asks, with
FORCE=1 for unattended use and a refusal rather than a hang when there
is no tty.
All four run clean; the kill path was tested against a live process and
left it alive.
jcoffey-dev
LINUXexpert.org