[GH-ISSUE #1845] Implementation of Enhanced Node Authorisation Features for Increased Security Across All User Tiers #3329

Open
opened 2026-08-05 00:51:52 -04:00 by saavagebueno · 8 comments
Owner

Originally created by @ezirius on GitHub (Apr 14, 2024).
Original GitHub issue: https://github.com/netbirdio/netbird/issues/1845

Is your feature request related to a problem? Please describe.
The current NetBird setup potentially allows insiders or compromised accounts to add unauthorised nodes to networks. This issue raises serious concerns not only for organisations but also for personal users who might store sensitive data or operate home-based IoT setups needing secure, controlled access. For instance, significant breaches in 2023, including those impacting large organisations like T-Mobile and Twitter, highlight the urgent need for enhanced security features to prevent unauthorised access​ (The Independent)​​ (NordLayer)​​ (IdentityIQ)​.

Describe the solution you'd like
I propose the introduction of a feature similar to Tailscale's "tailnet lock," but tailored to NetBird's system architecture. This feature would involve a robust node authorisation mechanism requiring explicit approval for each new node's integration into the network through a secure and verifiable process. Ideally, this would use cryptographic signatures verified against a list of pre-approved signatories within the user's network—ensuring no node joins without proper authorisation from a trusted entity. It's crucial that this feature operates independently of the central coordination server, which could be a vulnerability if compromised.

Describe alternatives you've considered
While a two-factor authentication system for adding new nodes, requiring secondary admin confirmation, could be an alternative, this system might still depend on the security of the central coordination server, posing a risk if compromised. This method could potentially be less secure than a decentralised signature verification system.

Additional context
The necessity for this feature goes beyond enterprise applications and is critical for personal users. The record-breaking number of breaches in 2023, along with the evolution of cyber threats from ransomware to data theft and extortion, shows the urgency of fortifying personal information. By implementing such security measures, NetBird would significantly enhance trust in its platform, ensuring that users at all levels—whether large enterprises or individual users—can maintain control over their networks without fear of internal threats or breaches. This feature is vital for upholding the integrity and security of private networks, protecting them against both external attacks and internal vulnerabilities (CRN)​.

Originally created by @ezirius on GitHub (Apr 14, 2024). Original GitHub issue: https://github.com/netbirdio/netbird/issues/1845 **Is your feature request related to a problem? Please describe.** The current NetBird setup potentially allows insiders or compromised accounts to add unauthorised nodes to networks. This issue raises serious concerns not only for organisations but also for personal users who might store sensitive data or operate home-based IoT setups needing secure, controlled access. For instance, significant breaches in 2023, including those impacting large organisations like T-Mobile and Twitter, highlight the urgent need for enhanced security features to prevent unauthorised access​ ([The Independent](https://www.independent.co.uk/advisor/vpn/live-data-breaches-and-stats))​​ ([NordLayer](https://nordlayer.com/blog/data-breaches-in-2023/))​​ ([IdentityIQ](https://www.identityiq.com/data-breaches/2023-a-year-of-record-breaking-data-breaches/))​. **Describe the solution you'd like** I propose the introduction of a feature similar to Tailscale's "tailnet lock," but tailored to NetBird's system architecture. This feature would involve a robust node authorisation mechanism requiring explicit approval for each new node's integration into the network through a secure and verifiable process. Ideally, this would use cryptographic signatures verified against a list of pre-approved signatories within the user's network—ensuring no node joins without proper authorisation from a trusted entity. It's crucial that this feature operates independently of the central coordination server, which could be a vulnerability if compromised. **Describe alternatives you've considered** While a two-factor authentication system for adding new nodes, requiring secondary admin confirmation, could be an alternative, this system might still depend on the security of the central coordination server, posing a risk if compromised. This method could potentially be less secure than a decentralised signature verification system. **Additional context** The necessity for this feature goes beyond enterprise applications and is critical for personal users. The record-breaking number of breaches in 2023, along with the evolution of cyber threats from ransomware to data theft and extortion, shows the urgency of fortifying personal information. By implementing such security measures, NetBird would significantly enhance trust in its platform, ensuring that users at all levels—whether large enterprises or individual users—can maintain control over their networks without fear of internal threats or breaches. This feature is vital for upholding the integrity and security of private networks, protecting them against both external attacks and internal vulnerabilities ([CRN](https://www.crn.com/news/security/the-10-biggest-data-breaches-of-2023-so-far))​.
saavagebueno added the feature-request label 2026-08-05 00:51:52 -04:00
Author
Owner

@nazarewk commented on GitHub (Apr 23, 2025):

@Ezirius
I might be downplaying the issue, but isn't the Peer Approval feature already covering this use case? It might not have been there yet when you originally posted the issue.

Image

Combining this with automated verification script/tooling created by and tailored to a specific organization's needs should be covering the whole use case?

<!-- gh-comment-id:2824504871 --> @nazarewk commented on GitHub (Apr 23, 2025): @Ezirius I might be downplaying the issue, but isn't the **Peer Approval** feature already covering this use case? It might not have been there yet when you originally posted the issue. ![Image](https://github.com/user-attachments/assets/f07bfbb9-ff8d-42f6-8863-3681cf25ab6f) Combining this with automated verification script/tooling created by and tailored to a specific organization's needs should be covering the whole use case?
Author
Owner

@YazeedAlKhalaf commented on GitHub (Jun 7, 2025):

@nazarewk the cool thing about the "Tailnet Locking" feature of Tailscale, is it doesn't rely on a central server, it is just public key cryptography that works and ensures security even if NetBird coordination central server is compromised, nobody can join your network unless a private key of a signing node signs the public key of the new node.

My details might not be 100% accurate but I think it covers the idea.

"Peer Approval" from what I understand is relying on NetBird to approve new nodes or people joining, right?

<!-- gh-comment-id:2952301780 --> @YazeedAlKhalaf commented on GitHub (Jun 7, 2025): @nazarewk the cool thing about the "Tailnet Locking" feature of Tailscale, is it doesn't rely on a central server, it is just public key cryptography that works and ensures security even if NetBird coordination central server is compromised, nobody can join your network unless a private key of a signing node signs the public key of the new node. My details might not be 100% accurate but I think it covers the idea. "Peer Approval" from what I understand is relying on NetBird to approve new nodes or people joining, right?
Author
Owner

@lixmal commented on GitHub (Jun 7, 2025):

Currently, you could set a PSK on each peer. The mgmt server won't know about it and peers that don't have it set won't be able to connect to the others

<!-- gh-comment-id:2952424349 --> @lixmal commented on GitHub (Jun 7, 2025): Currently, you could set a PSK on each peer. The mgmt server won't know about it and peers that don't have it set won't be able to connect to the others
Author
Owner

@dani commented on GitHub (Jul 27, 2025):

Unfortunatly, the peer approval feature is limited to netbird cloud, and not available in the self hosted edition. IMHO, this is a very essential security feature (even if indeed, a signature based feature like tailnet lock would be even better)

<!-- gh-comment-id:3124325911 --> @dani commented on GitHub (Jul 27, 2025): Unfortunatly, the peer approval feature is limited to netbird cloud, and not available in the self hosted edition. IMHO, this is a very essential security feature (even if indeed, a signature based feature like tailnet lock would be even better)
Author
Owner

@zaid-marji commented on GitHub (Feb 21, 2026):

I am a Tailscale user. And I chose Tailscale over Netbird specifically because of the Tailnet Lock feature. I would love to see this feature implemented in Netbird.

Tailscale has the equivalent of Peer Approval. The issue with this approach is that it does not protect against compromise of the coordination server as many others have pointed out.

A PSK is an improvement. However, the fatal flow of a PSK is that all users are now going to be trusted at the same highest level. And a PSK is shared. If we learned anything in the history of security, it is that sharing secrets is a huge vulnerability. Using a public cryptography approach that Tailnet Lock uses (where trusted devices digitally sign the public keys of the approved devices) is much safer than a PSK. A PSK also gives all users access to the PSK, which makes it much harder to control who has access to the secret.

So, while a PSK eliminates the coordination server as a source of vulnerability, it makes every device and every user in the network a potential vulnerability, while Tailnet Lock approach with public key cryptography ensures the private keys which are essential for the security of the system remain restricted to a few users and devices.

<!-- gh-comment-id:3938444882 --> @zaid-marji commented on GitHub (Feb 21, 2026): I am a Tailscale user. And I chose Tailscale over Netbird specifically because of the Tailnet Lock feature. I would love to see this feature implemented in Netbird. Tailscale has the equivalent of Peer Approval. The issue with this approach is that it does not protect against compromise of the coordination server as many others have pointed out. A PSK is an improvement. However, the fatal flow of a PSK is that all users are now going to be trusted at the same highest level. And a PSK is shared. If we learned anything in the history of security, it is that sharing secrets is a huge vulnerability. Using a public cryptography approach that Tailnet Lock uses (where trusted devices digitally sign the public keys of the approved devices) is much safer than a PSK. A PSK also gives all users access to the PSK, which makes it much harder to control who has access to the secret. So, while a PSK eliminates the coordination server as a source of vulnerability, it makes every device and every user in the network a potential vulnerability, while Tailnet Lock approach with public key cryptography ensures the private keys which are essential for the security of the system remain restricted to a few users and devices.
Author
Owner

@xerxist commented on GitHub (Apr 4, 2026):

Im surprised a simple feature which you can easily do with any reverse proxy takes priority of this... if the control plane is compromised the whole network is open... I'm staying with Tailscale for now

<!-- gh-comment-id:4187901494 --> @xerxist commented on GitHub (Apr 4, 2026): Im surprised a simple feature which you can easily do with any reverse proxy takes priority of this... if the control plane is compromised the whole network is open... I'm staying with Tailscale for now
Author
Owner

@lixmal commented on GitHub (Apr 5, 2026):

@xerxist what simple feature takes priority?

Pre-shared Keys will mostly achieve the same as the proposed solution

<!-- gh-comment-id:4188296787 --> @lixmal commented on GitHub (Apr 5, 2026): @xerxist what simple feature takes priority? Pre-shared Keys will mostly achieve the same as the proposed solution
Author
Owner

@xerxist commented on GitHub (Apr 5, 2026):

@xerxist what simple feature takes priority?

Pre-shared Keys will mostly achieve the proposed solution

That is not the same as node signing. Having signing nodes makes sure even if the coordination server is compromised you still can't communicate with the mesh network.

<!-- gh-comment-id:4188605272 --> @xerxist commented on GitHub (Apr 5, 2026): > @xerxist what simple feature takes priority? > > Pre-shared Keys will mostly achieve the proposed solution That is not the same as node signing. Having signing nodes makes sure even if the coordination server is compromised you still can't communicate with the mesh network.
Sign in to join this conversation.
No Label feature-request
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: DYNR/netbird#3329