I build a small, recoverable lab before I build a cluster. A working restore matters more to me than a long list of running services.
Before I start
- A spare x86-64 computer with virtualisation support and a wired network connection.
- My practical starting point is 16–32 GB RAM and an SSD; workload needs vary.
- A backed-up installation disk, a reserved management IP, and backup storage on a different device.
Step by step
Draw the minimum network plan
I record the host name, IP, gateway, DNS, and backup location before installation. Example: pve01.lab.home.arpa, 192.168.10.10/24, gateway 192.168.10.1. I keep management reachable from a trusted LAN or VPN, with no router forwarding to port 8006.
Download and boot the official installer
Download the Proxmox VE ISO from the vendor, verify the published checksum, and write it to USB. Enable CPU virtualisation in firmware, then boot the installer.
Install on the correct disk
The installer erases the selected target. I check the disk serial and size against my plan before proceeding. Set the region, a strong root password, a real notification address, host name, and network details.
Open the management interface
After reboot, open https://<management-IP>:8006 from the trusted LAN. Verify that the displayed host and certificate fingerprint belong to your own host before trusting a self-signed certificate. Check the node summary, network bridge, and local storage.
Update through the intended repository
Check the official repository instructions for the installed release. Use the enterprise repository with a subscription, or the documented no-subscription repository for this lab. I apply updates and reboot when required before loading important workloads.
Create one Linux VM
Upload an official Debian ISO. Create VM 101 with a modest starting allocation: 2 vCPU, 2–4 GB RAM, 32 GB disk, and the LAN bridge. Install Debian, apply updates, and record its IP. I avoid overallocating RAM before I understand the host’s capacity.
Back up to a separate destination
Add a supported backup destination in Datacenter > Storage, then create a backup job under Datacenter > Backup. Select the VM, a schedule, retention, and notifications. Run a first backup and inspect its task result.
Restore into an isolated test VM
Restore the backup under a new VM ID with its network disconnected to avoid duplicate addresses. Boot it, confirm a test file and service, then record elapsed restore time. I do not treat a snapshot on the same disk as a backup.
How I check the result
0 of 4 checks completed
Checklist progress stays in this browser. I use it to track checks, not to certify a deployment or compliance.
If something does not work
No interface after reboot
Check physical link, management IP, gateway, and bridge configuration on the console.
VM has no network
Check the selected bridge, VLAN assumptions, guest NIC, and DHCP reservation.
Backup failed
Read the task log and check destination reachability, free space, and permissions.
My working templates
I compare an actual recovery with the recovery-time and recovery-point objectives.
Download CSV ↓Official sources I checked
I checked these sources on 13 October 2026. Product screens and legal requirements change; I verify the relevant source before a production rollout or legal decision.