Rebooting or force-stopping your VPS

Use a graceful reboot first. From the shell on AlmaLinux 9 or Ubuntu 24.04, tell systemd to restart the VPS cleanly. Or use Restart in Hostworld’s Virtualizor panel. If the VPS is not responding to SSH or the network, open the Virtualizor VNC console and try a clean shutdown from there. Only force power off when graceful options fail because it risks data loss.

Before you start

• Know where to act. VPS power controls are in Virtualizor, reached from the Hostworld client area. Billing and account tasks are in WHMCS. If you need help from us, please open a support ticket.
• Have console access ready. If networking drops, you can still reach the server through Virtualizor’s VNC console, labelled “Launch HTML 5 VNC Client”.
• Avoid reboots during package operations. Rebooting while apt, dpkg or dnf is mid‑transaction can leave the OS inconsistent or unbootable. Wait until updates finish and prompts are cleared.
• Prefer graceful methods. Virtualizor’s Stop sends a shutdown signal to the guest. Restart restarts the VPS when the OS can still shut down cleanly. Poweroff immediately cuts power. Forced power cuts can corrupt filesystems or databases.
• Expect service restarts on Ubuntu 24.04. The needrestart tool restarts affected services after updates by default, so a full reboot is needed less often. Kernel or low‑level updates can still require a reboot.
• Decide if a reboot is needed. On Ubuntu, the file /run/reboot-required is created when a reboot is required. On AlmaLinux, the needs-restarting -r tool can report if a reboot is required.
• Network changes can lock you out. Firewall and interface changes can drop SSH on return. Validate your configuration before rebooting and keep the VNC console open during changes.
• Panel‑initiated graceful shutdowns are more reliable when the QEMU guest agent is installed and enabled inside the VPS. Without it, the guest may ignore shutdown signals.

Step 1: Restart from the Hostworld panel

Start with our panel controls if you can reach them. They integrate with the virtualisation layer and are available even if SSH is down.

• Use Restart for a normal reboot. This is the right choice when the OS is responsive enough to shut down cleanly. Virtualizor will report that the VPS was restarted successfully.
• Use Stop to send a shutdown signal to the guest when you want it to power down before you press Start again. This asks the OS to shut down. It is not an electrical cut.
• Avoid Poweroff unless nothing else works. Poweroff forces the VPS off immediately. Libvirt will report the domain as destroyed. This is equivalent to pulling the plug and can corrupt filesystems.
• Keep the VNC console handy. If you have changed networking or the firewall, launch the Virtualizor VNC first so you can log in on the console if SSH does not come back.

If you prefer a step‑by‑step walkthrough of our panel actions, use the VNC console to confirm the guest is actually booted, try Restart, fall back to Stop then Start, and reserve Poweroff for last resort.

Step 2: Reboot from the shell for a clean restart

This tells systemd to queue a normal shutdown then restart via reboot.target. It stops services cleanly, unmounts filesystems and reboots.

AlmaLinux 9

systemctl reboot

Ubuntu 24.04

systemctl reboot

The SSH session will end as the machine goes down. That is expected.

Forced flags bypass normal shutdown steps and risk data loss. Only use them in emergencies:

AlmaLinux 9

systemctl reboot -ff

This forces an immediate reboot through systemd, skipping service stop and unmount sequences. Use only if the system will not shut down any other way.

Ubuntu 24.04

reboot -f

On Ubuntu 24.04, reboot -f maps to systemctl reboot -ff. It bypasses normal shutdown and should be used only in emergencies.

Do not run any reboot command while package management is in progress. Wait for apt, dpkg or dnf to finish.

Step 3: Check if a reboot is actually required

There is no benefit to rebooting more often than needed. Check first so you can plan your window and avoid surprises.

Ubuntu 24.04

• When updates require a reboot, Ubuntu creates the file /run/reboot-required. Its presence means a reboot is needed.
• Ubuntu 24.04 restarts many services automatically after updates using needrestart. Kernel or core library updates can still require a full reboot.

AlmaLinux 9

This asks DNF’s plugin to report whether a restart is required after updates.

needs-restarting -r

The command returns a result indicating if a reboot is required. It is provided by dnf-plugins-core.

Step 4: Use the Virtualizor VNC console when locked out

When SSH or networking is down, use the VPS console to regain control.

1. From your VPS page in Virtualizor, launch the VNC console. The button is labelled “Launch HTML 5 VNC Client”.
2. Log in on the console and check whether the system is booted and responsive.
3. If the OS is healthy, perform a clean shutdown or reboot from the console using the commands in Step 2.
4. If the OS is unresponsive on the console as well, escalate to the options in Step 7 and Step 8.

Step 5: Fix networking after a reboot

Network or firewall changes can leave you without SSH when the VPS returns. Virtualizor’s guidance is:

• Check your firewall and any bridging rules for mistakes.
• Use the VNC console to verify the VPS actually booted and to repair configuration if needed.
• If the network interface is not attached to the hypervisor bridge, stop and start the VPS from the panel to reattach it.
• When editing networking, keep the VNC console open so you can undo a change that locks you out.

Step 6: Investigate what happened on the previous boot

This shows systemd’s journal from the previous boot so you can see why the last shutdown or startup was problematic.

AlmaLinux 9

journalctl -b -1

Ubuntu 24.04

journalctl -b -1

Look for services that failed to stop, filesystems that were not cleanly unmounted, or storage and network errors. If you had to use a forced reboot or Poweroff, pay close attention to filesystem and database messages.

Step 7: Recover with Virtualizor Rescue Mode when the VPS will not boot

Rescue Mode boots a predefined rescue OS so you can mount the original disk and repair or recover files. Enable it from the Manage VPS page in Virtualizor. First‑time activation can take 3 to 5 minutes. Once in rescue, mount your system disk, fix configuration or copy data, then disable Rescue Mode and boot the VPS normally.

If you are not sure how to proceed in Rescue Mode, please open a support ticket so we can look with you.

Step 8: Use emergency‑only options when you must

If the VPS will not shut down gracefully and you have confirmed on the VNC console that it is stuck, you may need to force a reboot or power cut. These actions risk data loss. Use them only after graceful methods fail.

• Forced reboot via systemd. This forces an immediate reboot and skips normal service stop and unmount work.
  • AlmaLinux 9:

    systemctl reboot -ff
    

  • Ubuntu 24.04:

    reboot -f
    

• Kernel Magic SysRq reboot. This triggers an immediate reboot in the kernel without syncing or unmounting filesystems. It is for dire emergencies only and carries a real risk of corruption on disk:

  echo b > /proc/sysrq-trigger
  

• Virtualizor Poweroff. This is equivalent to pulling the power. Libvirt reports the domain as destroyed. Use only when the OS is totally unresponsive on the console and be prepared to repair the filesystem afterwards.

Step 9: Make future panel‑initiated reboots more reliable

• Install and enable the QEMU guest agent inside the VPS. With the agent present, graceful shutdowns and restarts initiated from Virtualizor are more reliable. Without it, the panel relies on ACPI events that some guests ignore.
• Prefer OS‑level reboots when changing system components. If you are updating packages or altering critical services, run the reboot from the shell so systemd can stop everything in the right order.
• Leave the VNC console open during firewall or network changes so you have a way back in if SSH does not return.

What next

If you want to dig deeper into managing a Linux VPS, explore our VPS guides. If you are planning more capacity or a second server for maintenance windows, see our UK plans at Linux VPS.

If you have taken a forced action and the VPS still misbehaves, or you are not confident proceeding, please open a support ticket. Tickets go straight to our team and are tied to your account.