published priority 10 25 min. 40 XP.

Linux Service Management. systemctl, journalctl, sysvinit

Every NCAE service is managed by systemd. You will type `systemctl restart <service>` at least once per minute during competition. This lesson is the foundation. without it, nothing in the other lessons makes sense. It also teaches `journalctl`, which is where you find out why a service REFUSES to start.

DCWF roles: IT-411 Technical Support Specialist IT-451 System Administrator Services: SSH LoginSMB LoginDNS INT FWDPostgres AccessWWW Port 80WWW ContentWWW SSL

Objectives

Understand systemd as the init system (PID 1) and how it manages services
Use `systemctl` to start, stop, restart, reload, enable, disable, and check status
Use `journalctl` to read logs from any service with time filtering
Distinguish between services (.service), targets (.target), and timers (.timer)
Read a systemd unit file and identify ExecStart, Restart policy, and dependencies
Know where custom unit files live: /etc/systemd/system/ (admin) vs /lib/systemd/system/ (package)

Quick reference

Command	Purpose
systemctl status <svc>	Current status + last few log lines
systemctl start <svc>	Start the service now
systemctl stop <svc>	Stop it now
systemctl restart <svc>	Stop then start
systemctl reload <svc>	Apply config changes without stopping (if supported)
systemctl reload-or-restart <svc>	Prefer reload; fall back to restart
systemctl enable <svc>	Auto-start on boot
systemctl disable <svc>	Don't auto-start on boot
systemctl is-active <svc>	Exits 0 if active. useful in scripts
systemctl list-unit-files --type=service --state=enabled	All enabled services
systemctl list-timers --all	All systemd timers
systemctl cat <svc>	Print the unit file contents
systemctl daemon-reload	Re-read unit files after editing them
journalctl -u <svc>	All logs from a service
journalctl -u <svc> -n 50	Last 50 lines
journalctl -u <svc> -f	Follow in real time (tail -f)
journalctl -u <svc> --since '5 minutes ago'	Time-filtered
journalctl -p err --since today	All errors today, any unit

Common pitfalls

Editing a unit file and forgetting `systemctl daemon-reload`. systemd still uses the old version
`systemctl status` lies. it can show 'active (running)' briefly while the process dies; use `systemctl is-active` in loops
`journalctl -u <svc>` without `-n` dumps the entire history. can be megabytes
Confusing `enable` (auto-start on boot) with `start` (start now). they're independent
Using `service foo restart`. still works via sysvinit shim but systemctl is the native tool

How it works (walkthrough)

# A minimal systemd service unit file
# /etc/systemd/system/my-watchdog.service
[Unit]
Description=Keep the tarball restore loop alive
After=network.target

[Service]
Type=simple
ExecStart=/root/restore-loop.sh
Restart=always                 # if it dies, systemd restarts it
RestartSec=2

[Install]
WantedBy=multi-user.target      # start on boot at multi-user runlevel

# After writing the file:
#   systemctl daemon-reload         # pick up the new file
#   systemctl enable --now my-watchdog

Skill drills