reboot

Updated

September 4, 2026

Overview

reboot restarts the system. On modern systemd systems it is typically a symlink or wrapper around systemctl reboot. Prefer explicit systemctl forms in scripts for clarity and options (wall messages, force flags). Always ensure workloads can stop cleanly.

Syntax

reboot [options]
systemctl reboot
systemctl reboot --message="..."

Common Options

Option / form Description
(none) Normal reboot via init/systemd
--help Help
systemctl reboot -i Ignore inhibitors (careful)
systemctl reboot --force Forceful (see man)
systemctl reboot --firmware-setup Reboot to firmware setup (UEFI) when supported
shutdown -r now Equivalent classic form
shutdown -r +15 "msg" Scheduled reboot

Examples with Explanations

Normal reboot

sudo reboot
sudo systemctl reboot
sudo shutdown -r now

Scheduled

sudo shutdown -r +30 "Reboot for kernel update"
sudo shutdown -c                     # cancel

With broadcast message

sudo systemctl reboot --message="Kernel update; back in 5 minutes"

Inhibitors

systemd-inhibit --list
sudo systemctl reboot -i             # ignore inhibitors if stuck

Firmware setup (UEFI)

sudo systemctl reboot --firmware-setup

After kernel install

# apply new kernel
sudo reboot
uname -r                             # after boot

Containers / hosts

# inside a container, reboot may be blocked or reboot the host depending on privileges —
# almost always wrong. Restart the container instead:
podman restart myctr
systemctl restart myapp.service

Notes / Pitfalls

  • Unsaved work dies; notify users (wall, tickets) on multi-user systems.
  • Force flags can skip clean unmounts — risk filesystem recovery work.
  • Cloud VMs: reboot vs stop/start may change ephemeral networking or public IPs depending on provider.
  • Serial/console access ready before remote reboot of network-critical hosts.
  • reboot during package transactions can leave broken state — finish apt/dnf first.

2026-relevant notes

  • Live kernel patching reduces some reboot urgency; still needed for many updates.
  • Immutable OS / image-based systems reboot into new deployments more often — treat as normal.
  • Prefer orchestration drain (k8s) before node reboot.

Additional Resources

  • man reboot, man systemctl, man shutdown