The Proxmox Web UI Hardening
Your hypervisor's control panel closed to every address but the ones you name and locked behind a second factor, with a way back in we prove works from your console before we hand anything over.
Most Proxmox boxes arrive from the data centre with the control panel answering the whole internet, a certificate your browser refuses to trust, and one password that everybody shares. Whoever guesses that password owns every machine you will ever run on it.
We close the panel down to the addresses you name, put a second factor on every admin login, and give each person an account with the powers their job needs instead of shared root. Your own certificate goes onto the hypervisor itself, so your staff stop clicking through a warning to reach their own servers.
Locking down a hypervisor is the one job where doing it well and locking you out of your own hardware look identical from the outside. So we shut the door, then open it again from your console in front of you, before anything is signed off. Our engineers have been running remote infrastructure since 2003.
What's included
- Your control panel answering only the addresses you name, proved from both sides while you watch: your office gets in, and a connection from anywhere else finds nothing listening
- A second factor on every admin login, with printed recovery codes in your hands and one of them deliberately spent, so you have seen the back door open before you need it
- Named accounts holding only the powers each person needs, and the root password everybody currently shares retired into a break-glass envelope
- Your own certificate installed on the hypervisor, with its expiry date written into your handover so the renewal is a diary entry rather than an outage
- A way back in that has been used rather than assumed: your console reached, the lockdown undone from it, and the one-line undo written down for the night you need it
- Fewer ways in altogether, with the panel listening on your management network instead of every wire in the box
- Before-and-after evidence for every change, a plain-English handover of who can reach what, and a snapshot you can roll back to
How it works
- 01
Book and send access
Book online with your ABN or NZBN. We tell you exactly what access and material we need; the clock starts once we have it.
- 02
We do the work
An experienced employee who specialises in this task works through it, and asks before anything outside the scope.
- 03
Delivered with evidence
You get everything listed on this page, with a written account of what was done.
Before you order
Have these ready so we can start the moment you order. The delivery clock only runs while we have them.
- Root or sudo SSH access to each node (up to three), plus a web interface login for the first pass
- A way to reach each node when the network is down: your provider's console, IPMI, or someone at the machine
- The fixed addresses that should keep access: your office, your VPN exit, anywhere your admins work from
- Up to three named admins with the role each one needs, and the person who will keep the break-glass root password
- Your certificate, key and chain, if you want the browser warning gone
- Two short slots in your diary: one to narrow access, and one where each admin joins with their phone to set up a second factor
Who it's for
A business running Proxmox on a data-centre-delivered box where the admin panel is still reachable from anywhere and the root password is the one everybody uses.
Questions
How fast is delivery?
Standard delivery is within 30 weekday hours (≈ 2 working days). Express delivery — within 18 weekday hours (≈ 2 working days) — costs 25% more. The clock pauses while we wait on access, on the short window you pick for narrowing access, or on your admins being free to set up their second factor, so scheduling those around your day never counts against the deadline.
Could this lock us out of our own servers?
That is the risk this service is built around, so we deal with it first rather than last. Before anything is narrowed we confirm you can reach each node through its console, and a second session stays open the whole way through. Access rules go in before any blocking rule, never the other way round. Then we shut the door, reopen it from your console to prove the recovery path works, and hand you the one-line undo in writing.
Do you put a reverse proxy in front of it, and do you supply the certificate?
No proxy is needed. Proxmox takes a certificate directly, which is where yours goes, so you get the padlock without another moving part between you and your hypervisor. You supply the certificate, key and chain; we never buy, issue, renew or automate one, because a renewal we set up is one that stops working a year later without anybody noticing. If you do want a proxy in front for header rules or routing, that is The Reverse Proxy Setup.
Is this the same as hardening the server itself?
No. This is the management plane: who can reach the control panel, who they have to prove they are, and what they can do once inside. The Linux underneath it — SSH policy, kernel settings, password rules, audit logging, the host firewall — is The Linux Server Hardening, and automatic banning of repeated login attempts is The Brute Force Guardian. The three sit on top of each other, and plenty of customers book them together.
Will you switch us to the no-subscription repository and get rid of the nag screen?
We record which repository each node uses and whether it can fetch updates at all, and you get the exact one-line fix if it cannot. We stop there deliberately. Which repository you pull from is a licensing and update decision rather than a security one, and it belongs with The Proxmox Audit & Tune-up, which covers your whole update path. The nag screen we leave alone: removing it means editing a file Proxmox replaces on every upgrade.
Our admins work from home on addresses that change. Can you still restrict access?
Not safely, and we would rather say so before you book than lock you out afterwards. Restricting by address needs an address that stays put — an office line, a VPN exit, a fixed range. If you have none, that item is recorded as blocked and everything else still runs. The usual fix is to give yourself one fixed way in first, which is The Zero Trust Network Setup, and then this service has something solid to lock to.
Often booked alongside
- DevOps & Servers
The Brute Force Guardian
Repeated failed logins get the attacker's IP banned at your firewall automatically — a standalone fail2ban setup, with your own addresses whitelisted first so you're never locked out.
A$135 From A$67.50 on a plan
- DevOps & Servers
The CI Runner & Pipeline Setup
Build machines wired into the git platform you already run, with up to three pipelines handed over running green on your own code.
A$354 From A$177 on a plan
- Proxmox Administration
The Cloud-Init Template Library
Four tested starting points for your virtual machines, so a new server is a two-minute clone rather than another operating-system install, and every clone comes up with an identity of its own.
A$249 From A$124.50 on a plan