Recommended Free Tools
If you ran sudo shutdown -h now, you powered the VPS off rather than rebooting it. Wait a moment for the shutdown to finish, then open your VPS provider’s control panel and choose Start, Power on, or the equivalent action for a stopped instance. SSH cannot start a guest that is already powered off. If you still have a live console before shutdown completes, use sudo shutdown -c only if the shutdown was scheduled for later; now means immediate execution, so cancellation may arrive too late.
Antihost offers Linux VPS plans with full root access, unlimited traffic on a 1 Gbps port and a 14-day free trial.
What shutdown -h now actually does
On a systemd-based Linux VPS, shutdown schedules a system shutdown and sends the selected action to the init system. The now argument is an alias for +0, meaning there is no delay. The -h option selects halt or poweroff behavior; unless an explicit halt action is requested, it is treated as --poweroff. In practical VPS terms, the guest stops running and its virtual power state changes to off.
As an Amazon Associate I earn from qualifying purchases.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
This is different from a reboot. A reboot shuts services down and then starts the operating system again. A poweroff ends the guest at the shutdown stage, so network services, SSH, and your shell disappear. The provider’s host still exists, but the virtual machine needs a start or power-on operation from outside the guest.
Recovery: turn the stopped VPS back on
- Wait briefly. Give the operating system time to flush data and finish stopping services. Testing SSH repeatedly during this interval does not accelerate the process.
- Test SSH once. If the session has already closed or the connection times out, treat that as confirmation that the guest is no longer accepting connections.
- Open the VPS provider dashboard. Sign in through the provider’s normal web control plane, not through the unavailable guest.
- Select the affected instance. Check its reported state. Look for labels such as Stopped, Powered off, Shut down, or Terminated. A state of Running may indicate that the shutdown is still progressing or that the provider has not refreshed its status.
- Choose Start or Power on. Use the action intended for a stopped instance. Confirm any warning dialog and wait for the state to change to Running.
- Open the browser, serial, or HTML console if offered. Watch the boot messages to confirm that the kernel starts, filesystems mount, and the login prompt appears. A console is especially useful when SSH remains unavailable after the control-panel action.
- Reconnect with SSH. Use the same address, port, user, and key as before. A refused connection usually means the machine is booting but the SSH service is not ready; a timeout can indicate that networking has not started or that the provider action did not complete.
- Check the guest after login. Verify mounts, failed services, application status, and recent boot messages before treating the recovery as complete.
Start, reboot, and reset are not interchangeable
Control panels commonly expose several power controls. Start or Power on is the normal choice when the instance is stopped. Reboot is generally intended for a running instance and asks the guest or virtualization layer to restart it. Reset, Hard reboot, or a similarly named action may cut power and start it again without allowing a clean shutdown.
Provider semantics differ. Some APIs define restart as a stop-then-start hard power cycle rather than a graceful in-guest reboot. Read the action’s description and check the reported power state before confirming it. If the dashboard says the VPS is stopped, use Start first; do not assume that a button labeled Restart will perform a graceful reboot.
If the shutdown has not finished yet
shutdown -c cancels a shutdown that is scheduled for a future time. It is useful after commands such as shutdown -h +10, where the system has a ten-minute delay. With shutdown -h now, execution begins immediately. Services may already be stopping and the command may have passed the point at which cancellation is reliable.
If a console or SSH shell is still responsive, you can try:
sudo shutdown -cThen check whether the cancellation message is accepted and whether the system remains operational. Do not run another power command simply because the prompt is slow to return; the original shutdown may still be working.
The right command when you meant to reboot
Use an explicit reboot action whenever the goal is to restart the operating system and return to the login prompt automatically.
| Intent | Command | Effect and cautions |
|---|---|---|
| Immediate poweroff | sudo shutdown -h now |
Selects immediate halt/poweroff. The VPS becomes unavailable until it is started externally. |
| Graceful reboot | sudo shutdown -r now |
Selects reboot, allowing the normal shutdown sequence before starting Linux again. |
| systemd reboot | sudo systemctl reboot |
Requests a reboot directly through systemd from a live shell. |
| Forced reboot | sudo reboot -f or the equivalent forced systemd form |
Bypasses parts of the init-system shutdown path. Use only when graceful methods fail; filesystems may not be properly unmounted. |
The -r or --reboot option is the important difference from -h. Do not substitute a forced reboot for an ordinary restart just to save a few seconds. Skipping service stop jobs and filesystem unmounts can leave data buffered or metadata inconsistent.
What to verify after the VPS starts
A green provider status does not prove that every guest service recovered. Perform checks in a deliberate order so that an application failure is not mistaken for a power or networking problem.
Confirm the operating system and boot time
uptime
who -b
systemctl is-system-runninguptime and who -b show whether a new boot occurred. systemctl is-system-running reports the systemd state; a result other than running deserves investigation, although a transitional state can be normal immediately after login.
Check failed units
systemctl --failedReview every listed unit rather than restarting everything blindly. For a particular service, inspect its state and recent messages:
systemctl status SERVICE_NAME
journalctl -u SERVICE_NAME -b --no-pagerReplace SERVICE_NAME with the actual unit, such as an application, web server, database, or SSH service. The -b filter limits journal output to the current boot, which helps separate a new failure from older incidents.
Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Clear out junk files and repair common Windows errors3Fix the driver behind crashes, sound loss and screen glitchesVerify disks and mounts
findmnt
df -h
lsblkLook for missing data volumes, read-only mounts, unexpectedly full filesystems, or a device that did not appear. Applications that rely on a separate volume can start unsuccessfully even though the operating system itself booted normally.
Check networking and SSH
ip addr
ip route
systemctl status sshConfirm that the expected interface has an address, a default route exists, and the SSH service is active. If the provider assigned a dynamic address, verify the current address in the dashboard before troubleshooting keys or firewall rules.
Inspect the current boot log
journalctl -b -p warning..alert --no-pagerThis displays warning-level and more serious messages from the current boot. Pay attention to storage, filesystem, network, and service errors that coincide with the failed recovery.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteWhen the dashboard action fails
The instance remains stopped
Refresh the instance page and check whether the provider reports an operation in progress. If the Start or Power on request fails repeatedly, use the provider’s documented API or support channel. Record the instance identifier, the displayed power state, the time of the failed request, and any operation ID or error text before opening a ticket.
The instance says running but SSH times out
Use the provider’s browser or serial console. If the console shows a boot prompt, wait for networking and SSH to finish. If it shows a filesystem check, emergency mode, or a kernel error, resolve that condition from the console rather than issuing repeated power cycles.
The instance is running but connections are refused
A refusal normally means that the address is reachable but no process is listening on the requested port. From the console, check systemctl status ssh, firewall rules, and the configured SSH port. Review journalctl -b for a failed network or SSH unit.
The provider offers only Reset
Use a hard reset only when the guest is genuinely stuck and graceful controls do not work. A reset can interrupt writes and skip clean unmounting, so after the machine returns, inspect filesystems, mounts, and service logs before resuming normal workloads.
Free tools Windows power users keep installed
One-click scans. No signup required.
Why SSH cannot be used to power on the guest
SSH is an application running inside the VPS. Once the guest is powered off, its virtual CPU is stopped, its network stack is inactive, and the SSH daemon cannot accept a command. No SSH option can start a machine that is not running. The start request must come from the provider’s control plane, a provider API, or an out-of-band console associated with the host.
A browser or serial console is also useful when the guest is powered on but unreachable over the network. It provides visibility into boot output without depending on the VPS’s interface, route, firewall, or SSH configuration.
Quick Recap
Safer operating habits for future maintenance
- Use
sudo shutdown -r noworsudo systemctl rebootwhen you intend to restart Linux. - Use
sudo shutdown -h nowonly when you deliberately want the VPS powered off and know how to start it from the provider side. - For a delayed action, specify the delay clearly, for example
sudo shutdown -r +5, and remember thatshutdown -ccan cancel a future schedule. - Keep the provider dashboard, API credentials, and console procedure available before starting maintenance that could end your SSH session.
- Before a reboot, confirm that important writes, database operations, and deployment jobs have completed.
- After every maintenance reboot, check failed units, mounts, networking, and application health instead of relying only on the provider’s Running label.
Quick decision guide
- You ran
shutdown -h nowand SSH is gone: wait briefly, then use Start or Power on in the provider dashboard. - The dashboard says the VPS is still running: use the provider console to watch shutdown or choose the documented reboot action.
- The command was scheduled for later: run
sudo shutdown -cwhile the shell is still available. - You wanted a reboot: next time use
sudo shutdown -r noworsudo systemctl reboot. - The VPS will not start: collect the displayed state and error details, then use the provider’s API or support process.
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
