Your Fast Hive VPS can be scaled, restarted, snapshotted and rebuilt from the client area. These are powerful operations and some of them are irreversible, so this guide explains exactly what each one does, when to reach for it, and the order to try them in when something has gone wrong.
The operations, at a glance
| Operation | What happens | Data loss? | Typical downtime |
|---|---|---|---|
| Reboot | Graceful restart, as if you ran reboot |
No | Under a minute |
| Power off / on | Hard stop, equivalent to pulling the plug | Unwritten data may be lost | Until you power it back on |
| Snapshot | Captures the whole disk at a point in time | No | None to brief |
| Restore snapshot | Rolls the disk back to that moment | Everything since the snapshot is lost | Minutes |
| Rebuild / Reinstall | Wipes the disk, installs a fresh OS | Total. Everything is erased. | About a minute to deploy |
| Resize / Scale | Changes vCPU, RAM or disk allocation | No | One reboot |
Where to find these controls
- Sign in at my.fasthive.com.
- Open My Products & Services and click your VPS.
- The service page shows the server's status, IP address and resource allocation, with the management actions alongside.
Taking and using snapshots
Snapshots are included on every VPS tier. Treat them as an undo button for changes you are about to make.
When to take one
- Immediately after you finish configuring a new server, as a known-good baseline.
- Before a distribution upgrade or a major package change.
- Before a deployment you are not confident about.
- Before editing anything that could lock you out - SSH configuration, firewall rules, authentication.
Taking one
- Open the VPS service page in the client area.
- Choose Snapshots.
- Click to create a new snapshot and give it a name that says why you took it, such as
before-php84-upgrade. In six weeks, "snapshot 3" tells you nothing. - Wait for it to complete before starting the change it protects.
Restoring one
- Select the snapshot and choose Restore.
- Read the confirmation carefully: everything written since that snapshot will be gone.
- If there is any data you need from the current state - a database, uploaded files, logs - copy it off the server first.
- Confirm, and wait for the server to come back up.
Rebooting safely
Always try a graceful Reboot before a hard power cycle. If SSH still works, rebooting from inside the server is gentler still:
sudo systemctl rebootUse the client area's power controls when SSH is unreachable. A hard power off is a last resort - it risks filesystem damage in exactly the way an unexpected power cut would.
Rebuilding the server
A rebuild reinstalls a clean operating system. Fast Hive offers Ubuntu 24.04, Debian 12, AlmaLinux 9, Fedora 41 and CentOS Stream, and a rebuild deploys in around 55 seconds.
Rebuild when you want a clean start, when you want to switch distribution, or when a server has been compromised and cannot be trusted.
- Copies of all website files, application code and configuration.
- Database dumps, downloaded and verified.
- Any certificates and private keys you cannot reissue.
- A note of installed packages and custom configuration.
- Your SSH public key ready to supply during deployment.
After a rebuild the host key changes, so SSH will warn you. Clear the stale entry:
ssh-keygen -R your.vps.ip.addressScaling CPU, memory and disk
Fast Hive VPS plans are sized by resources rather than feature gates, and CPU and memory can be scaled later from the panel with billing following by the hour rather than the year.
- Open the VPS service page and choose the upgrade or change-plan option.
- Select the new size and review the pricing.
- Confirm. The server reboots into the new allocation.
Before scaling up, check that you are genuinely resource-bound rather than misconfigured:
free -h # memory, including swap
df -h # disk usage
uptime # load averages
top -o %CPU # what is actually consuming CPUSustained load average above your core count, or memory pressure forcing constant swapping, is a real case for more resources. A single runaway process is not.
Troubleshooting: what to try, in order
| Situation | Try this order |
|---|---|
| Website down, SSH works | Check the web server's status and logs. Do not reboot - you would lose the evidence. |
| SSH unreachable, site still up | Use console access from the client area. Check the firewall and the SSH service before rebooting. |
| Everything unreachable | Console access, then graceful reboot, then hard power cycle, then open a ticket. |
| Locked out by your own config change | Console access to undo it. Restore a snapshot only if that fails. |
| Disk full | Find the culprit with du -sh /* | sort -h, clear logs and package caches, then consider resizing. |
| Server will not boot after a kernel update | Use console access to select the previous kernel at the boot menu, then restore a snapshot if needed. |
| Suspected compromise | Do not reboot or rebuild yet. Open a ticket immediately so the entry point can be identified first. |
Frequently asked questions
Do snapshots count as backups?
They are a genuine safety net — nightly images are taken automatically and kept off-server, so they survive a problem with the server itself. What they are not is a substitute for application-level backups: a disk image restores the whole machine to a point in time, which is a blunt instrument if you only need last night's database. Keep your own database dumps as well.
How long does a snapshot take?
Usually a few minutes, depending on how much data is on the disk. Restores are similar.
Will a reboot change my IP address?
No. Your IP address is assigned to the service and stays with it across reboots, snapshot restores and rebuilds.
Can I scale back down after scaling up?
CPU and memory can be scaled in both directions. Disk is normally grow-only, since shrinking a filesystem risks data. Check the options on your service page or ask Sales & Solutions.
Does scaling cost a full month?
No. VPS billing follows by the hour, so a temporary increase for a busy period costs only for the hours you use it.
What happens to my data when I cancel a VPS?
It is removed. Take a full backup before cancelling - there is no recovery afterwards.
