Introduction
	There are two different operations people often call “restarting the server,” and it is worth keeping them apart. Most of the time you only need to restart a single service — for example, the web server after a configuration change — which takes a second and affects nothing else. Far less often you need to reboot the whole VPS, which restarts the operating system and briefly takes everything offline. This page covers both, for FreeBSD 15 and Rocky Linux 10. Day-to-day service management lives on the companion Server Maintenance page.
	Restarting a Service
	After you edit a service’s configuration, the change takes effect only once the service re-reads it. Unlike older VPS platforms, editing a config file does not restart anything automatically — you do it yourself. Two verbs matter: reload re-reads the configuration gracefully, without dropping active connections (use this whenever a service supports it), while restart fully stops and starts the service. Using the web server as the example, always test the configuration first so a typo cannot take the site down: # FreeBSD # apachectl configtest # check the config before applying it # service apache24 reload # apply changes, keep connections alive # service apache24 restart # full stop then start # Rocky Linux # apachectl configtest # systemctl reload httpd # systemctl restart httpd The same pattern works for any service — substitute its name (sendmail, dovecot, mariadb, proftpd, tomcat, and so on): service <name> restart on FreeBSD, systemctl restart <name> on Rocky Linux. Restarting one service leaves the rest of the server untouched, so prefer it to a full reboot whenever it will do.
	Rebooting the Whole VPS
	A full reboot is rarely necessary — chiefly after a kernel update (a new kernel from freebsd-update on FreeBSD, or a dnf kernel update on Rocky Linux), since a running kernel can only be replaced by restarting. Almost everything else is handled by restarting the relevant service. When you do need to reboot, run one of these as root: # shutdown -r now # reboot immediately # reboot # the same thing # shutdown -r +5 "rebooting for kernel update" # in 5 minutes, warning logged-in users Both commands work identically on FreeBSD and Rocky Linux. The server will be unreachable for a minute or so and your SSH session will drop; reconnect once it is back. TIP: Before rebooting, make sure every service you rely on is enabled to start at boot, or it will not come back on its own: sysrc <name>_enable=YES on FreeBSD, systemctl enable <name> on Rocky Linux. After a kernel update, schedule the reboot (the +5 form above) for a quiet moment rather than the middle of your busiest hour.
	Shutting Down
	To power the machine off rather than reboot it, use shutdown -p now on FreeBSD or poweroff (equivalently shutdown -h now) on Rocky Linux. WARNING: On a remote VPS you usually cannot power the machine back on yourself — a power-off leaves it off until it is started again for you. Use reboot when you want the server to come back; only power off if you intend it to stay off, and contact support if you need a full power cycle you cannot perform yourself.
	Confirming the Server Is Back
	After a reboot, check that the system and your services came back up: $ uptime # a low uptime confirms it just rebooted $ last reboot # the history of reboots # Rocky Linux: has the boot finished cleanly? # systemctl is-system-running # prints "running" when all is well # FreeBSD: what is set to start at boot? # service -e Finally, confirm the service itself is healthy — service apache24 status or systemctl status httpd — and load your site in a browser. If something did not return, start it by hand and check its logs for why.
	Documentation
	See the shutdown(8) and reboot(8) manual pages for every option. Patching and kernel updates — the usual reason to reboot — are covered on the Securing Your VPS page.