Skip to content

Contabo Resets Every VNC Password After a Break-In: What Server Customers Should Check

Published

Laura Bede

By Laura Bede

Journalist

Miklos Zoltan

Fact-Checked this

Contabo, one of Europe’s best-known budget server hosts, says someone got into its systems on 14 November.

The company posted a short incident notice on its help site at 9:00 pm CET on 16 November. Among its first steps, it reset the VNC passwords of all customers. VNC is the remote console that shows a server’s screen, so the reset puts the console at the centre of this incident, even though Contabo hasn’t said the attacker used it.

Contabo says it has “no indications that personal data was leaked” and that its services run as usual. When we published, the notice had reached Hacker News and Mastodon, and two small blogs in Germany and Italy had written it up. No large outlet had reported it.

If you rent a VPS or VDS from Contabo, the checklist is in chapter 3.

What Contabo Has Said

The notice is brief. According to Contabo, it detected unauthorized access on Saturday, 14 November, stopped it, and called in outside incident-response specialists for a forensic investigation. It lists three steps taken so far:

  • reset all VNC passwords and blocked “the known access paths”
  • rotated internal access credentials
  • added monitoring and further forensic measures

Its advice to customers is general: watch for suspicious activity, review credentials and access permissions, change default passwords, turn on multi-factor authentication and be careful with links in unexpected messages. Contabo says it may not answer individual questions about the incident and will post confirmed findings on the same page every day by 2:00 pm CET. The German site Techkrams reports that customers were also told by email.

Just as telling is what the notice leaves out:

  • how the attacker got in
  • which systems they reached
  • when the access began (14 November is the day Contabo noticed it)
  • whether the attacker saw or used any VNC passwords, or reached any customer server

Until Contabo answers these, nobody outside the company can say whether customer servers were at risk. The reset of every VNC password is the strongest hint about where Contabo is looking.

Why the VNC Reset Matters

On a Contabo VPS or VDS, VNC is the emergency door. In Contabo’s own words, it “connects directly to the server’s display”, so it works even when SSH is broken or the system won’t start properly. You connect with a VNC client to an address and port that Contabo gives you, and a single password lets you in. The Customer Panel shows that password on request, which suggests Contabo keeps it in a form it can read back. That last point is our reading of how the panel works, not something Contabo has said.

A VNC password on its own isn’t the server’s root password. Once connected, you still see a login prompt. Two things make the console more dangerous than that sounds.

First, sessions don’t close themselves. Contabo’s guide warns that “VNC sessions do not end automatically when you close the client window.” If you once logged in as root through VNC and only closed the window, the next person with the VNC password lands in that logged-in shell.

Second, the console is where a locked-out admin resets a forgotten root password. On many Linux servers, anyone who can watch the server restart can open the boot menu and change the root password in a few minutes, as Red Hat’s documentation describes step by step. A boot-loader password or full-disk encryption stops that, but both have to be set up on purpose.

None of this means an attacker did any of it. It explains why Contabo reset the passwords, and why the reset alone doesn’t settle the question for your server: if someone got in before 14 November, a new VNC password doesn’t undo what they did.

Contabo’s own guide already says VNC “is less secure than SSH” and should be switched off when you don’t need it. This week is a good time to take that advice.

What to Check Now

What to check
Where
Why
Turn VNC off if you don’t use it
Customer Panel → Servers & Hosting → VPS/VDS → More (three dots) → VNC Control
Contabo’s own guide calls VNC less secure than SSH and says to disable it when not in use.
Close console sessions left open
Connect once with the new VNC password, type logout (Linux) or sign out (Windows), then disable VNC
Closing the VNC window doesn’t end the session. A shell left logged in stays open for whoever connects next.
Logins since early November
Linux: last -F, lastb, journalctl --since 2026-11-01. Windows: Event Viewer → Security, event 4624 with logon type 2 (at the console)
Contabo gave the day it detected the access, not the day it began.
New users, SSH keys, cron jobs, services
/etc/passwd, every ~/.ssh/authorized_keys, crontab -l, systemctl list-timers, recently changed files in /etc/systemd/system
These are the usual ways to keep a way back in after a password change.
Customer Panel login
New password; two-factor login in the old and the new panel; new API credentials
Contabo’s help says 2FA must be set up in each panel separately.
Keys stored on the server
Database passwords, API tokens, WireGuard or other VPN server keys
Only if something above looks wrong: anyone at the console could read them.
Emails about the incident
Read updates on Contabo’s incident page, not through email links
Contabo says it publishes every confirmed finding there daily by 2:00 pm CET.
Red rows first. None of these steps tells you whether your server was touched; they close the doors the incident could have opened and show you where to look.
Sources: Contabo incident page (16 Nov 2026) and Contabo help articles on VNC and compromised accounts, checked 17 November 2026.

Start with the console. If you don’t use VNC, turn it off in the Customer Panel. If you do, log in once with the new password, make sure no shell is left open, and then switch it off until you need it.

Next, look for logins you don’t recognise from the start of November onward, not only from the 14th. On Linux, last -F lists successful logins and lastb failed ones; console logins show up as tty1 rather than an IP address. On Windows, interactive logons at the console are event 4624 with logon type 2.

Then look for ways back in that survive a password change: user accounts you didn’t create, unfamiliar lines in authorized_keys, new cron jobs, timers or services.

If you run your own VPN on a Contabo server, for example WireGuard or Amnezia, its private keys sit on that server. If anything in the checks above looks wrong, generate new server keys and send your devices new configs.

If Something Looks Wrong

Don’t try to clean a server you no longer trust. Back up your data, not the system, then reinstall from a fresh image and restore. Change every secret that lived on the old server: database passwords, API tokens, VPN keys, and any SSH keys that were stored there.

If the server held personal data of customers or staff, you’re the one responsible for it under the GDPR, not Contabo. A personal-data breach has to be reported to your data protection authority within 72 hours of you becoming aware of it, unless it’s unlikely to put people at risk. Our GDPR compliance checklist explains the steps.

For the Customer Panel itself, Contabo’s guide for compromised accounts is useful even if nothing happened to yours. It says two-factor login has to be set up separately in the old and the new Customer Panel, and it tells you how to create new API credentials.

What We're Asking Contabo

We’re contacting Contabo to ask how the attacker got in, when the access began, whether they reached the systems that hold VNC passwords or any customer server, and whether customers whose servers were touched will hear from Contabo directly.

We’ll add the answers here, along with anything Contabo’s daily updates confirm.

Our Mission

We believe security online security matters and its our mission to make it a safer place.