[GH-ISSUE #5535] Networks on MacOS don't appears to work at all #10721

Closed
opened 2026-08-05 01:27:02 -04:00 by saavagebueno · 4 comments
Owner

Originally created by @rihards-simanovics on GitHub (Mar 7, 2026).
Original GitHub issue: https://github.com/netbirdio/netbird/issues/5535

Describe the problem

In a recent couple of minor versions, I've noticed that macOS seems to completely disregard networks. I have some configuration which were allowing some internal services to be exposed via networks (before Reverse Proxy was a feature), now said services are completely unreachable.

To Reproduce

Steps to reproduce the behaviour:

  1. set up a server with a service hosting websites (at least 2 one public one behind IP block). Some websites on said server can be accessed publicly, others are only accessible by visiting through exit node.
  2. set up said exit node
  3. set up a client on macOS to use said exit node above to access the blocked site via the exit node.
  4. set up another client now Linux Ubuntu, and try accessing the blocked site via exit node.

The macOS should fail with a 403 and Linux Ubuntu should work.

Expected behaviour

macOS should use the networks to route the traffic for services accessible only through exit nodes

Are you using NetBird Cloud?

selfhosted

NetBird version

client, exit node, management server, are all running the v0.66.2

Is any other VPN software installed?

nope

Debug output

I will only after I can confirm that others don't have this issue

To help us resolve the problem, please attach the following anonymized status output

  netbird status -dA

Create and upload a debug bundle, and share the returned file key:

  netbird debug for 1m -AS -U

*Uploaded files are automatically deleted after 30 days.*


Alternatively, create the file only and attach it here manually:

  netbird debug for 1m -AS

Screenshots

Not applicable, I simply get 403 forbidden with nginx, as the IP of the connecting party is different from expected exit node one.

Additional context

Worked fine before, no configurations changed, only versions.

Have you tried these troubleshooting steps?

  • Reviewed client troubleshooting (if applicable)
  • Checked for newer NetBird versions
  • Searched for similar issues on GitHub (including closed ones) as good as useless issues don't get closed at all
  • Restarted the NetBird client
  • Disabled other VPN software
  • Checked firewall settings
Originally created by @rihards-simanovics on GitHub (Mar 7, 2026). Original GitHub issue: https://github.com/netbirdio/netbird/issues/5535 **Describe the problem** In a recent couple of minor versions, I've noticed that macOS seems to completely disregard networks. I have some configuration which were allowing some internal services to be exposed via networks (before Reverse Proxy was a feature), now said services are completely unreachable. **To Reproduce** Steps to reproduce the behaviour: 1. set up a server with a service hosting websites (at least 2 one public one behind IP block). Some websites on said server can be accessed publicly, others are only accessible by visiting through exit node. 2. set up said exit node 3. set up a client on macOS to use said exit node above to access the blocked site via the exit node. 3. set up another client now Linux Ubuntu, and try accessing the blocked site via exit node. The macOS should fail with a 403 and Linux Ubuntu should work. **Expected behaviour** macOS should use the networks to route the traffic for services accessible only through exit nodes **Are you using NetBird Cloud?** selfhosted **NetBird version** client, exit node, management server, are all running the `v0.66.2` **Is any other VPN software installed?** nope **Debug output** I will only after I can confirm that others don't have this issue ``` To help us resolve the problem, please attach the following anonymized status output netbird status -dA Create and upload a debug bundle, and share the returned file key: netbird debug for 1m -AS -U *Uploaded files are automatically deleted after 30 days.* Alternatively, create the file only and attach it here manually: netbird debug for 1m -AS ``` **Screenshots** Not applicable, I simply get 403 forbidden with nginx, as the IP of the connecting party is different from expected exit node one. **Additional context** Worked fine before, no configurations changed, only versions. **Have you tried these troubleshooting steps?** - [x] Reviewed [client troubleshooting](https://docs.netbird.io/how-to/troubleshooting-client) (if applicable) - [x] Checked for newer NetBird versions - [x] ~~Searched for similar issues on GitHub (including closed ones)~~ as good as useless issues don't get closed at all - [x] Restarted the NetBird client - [x] Disabled other VPN software - [x] Checked firewall settings
saavagebueno added the triage-needed label 2026-08-05 01:27:02 -04:00
Author
Owner

@rihards-simanovics commented on GitHub (Mar 19, 2026):

This is the kind of setup I have on Linux and macOS, both are identical in the number of networks they have and the kind of groups they are assigned to:

Image

This worked really well before and nothing changed in the setup. macOS however entirely ignores the network configurations in the management server despite all being enabled and still doesn't route through an exit node.

Image

I have a feeling that the client on the macOS doesn't update the routing table at all, I'll print out the network's tan as well as the routing table from macOS

<!-- gh-comment-id:4087974047 --> @rihards-simanovics commented on GitHub (Mar 19, 2026): This is the kind of setup I have on Linux and macOS, both are identical in the number of networks they have and the kind of groups they are assigned to: <img width="2802" height="1529" alt="Image" src="https://github.com/user-attachments/assets/3a6b1f38-90d5-4c23-9560-c6800765ec3f" /> This worked really well before and nothing changed in the setup. macOS however entirely ignores the network configurations in the management server despite all being enabled and still doesn't route through an exit node. <img width="2802" height="1529" alt="Image" src="https://github.com/user-attachments/assets/62680646-e5c3-44fe-bf64-d190460e142f" /> I have a feeling that the client on the macOS doesn't update the routing table at all, I'll print out the network's tan as well as the routing table from macOS
Author
Owner

@rihards-simanovics commented on GitHub (Apr 6, 2026):

@mlsmaycon sorry to ping directly, but I am still experiencing this issue.

<!-- gh-comment-id:4194679145 --> @rihards-simanovics commented on GitHub (Apr 6, 2026): @mlsmaycon sorry to ping directly, but I am still experiencing this issue.
Author
Owner

@jnfrati commented on GitHub (Apr 7, 2026):

@rihards-simanovics posted images don't seem to be working, could you provide the debug bundle from the template?

<!-- gh-comment-id:4198682781 --> @jnfrati commented on GitHub (Apr 7, 2026): @rihards-simanovics posted images don't seem to be working, could you provide the debug bundle from the template?
Author
Owner

@rihards-simanovics commented on GitHub (Apr 20, 2026):

@rihards-simanovics posted images don't seem to be working, could you provide the debug bundle from the template?

Hi @jnfrati, I managed to resolve the issue. Essentially, one of the policies that permitted the client to connect to the exit node (so that the said exit node can be used reached), was deleted. Now, I don't know what the underlying issue is, so you will have to tell me, but the symptoms are:

  1. Two clients have the same groups assigned to them.
  2. Group policy permitting access to exit node just, disappears -- this is important as once we set up the groups and policies we tend not to touch them.
  3. One client somehow still can connect to the exit node, the other can't.

I can only think of some migration problem on a DB side in the management server because this doesn't make sense. I've seen in the past how, a port number, for instance, applied to the access policy, but doesn't show in the UI.

I will close this as “fixed” but how it has broken in a first place I have no clue.

<!-- gh-comment-id:4284755627 --> @rihards-simanovics commented on GitHub (Apr 20, 2026): > [@rihards-simanovics](https://github.com/rihards-simanovics) posted images don't seem to be working, could you provide the debug bundle from the template? Hi @jnfrati, I managed to resolve the issue. Essentially, one of the policies that permitted the client to connect to the exit node (so that the said exit node can be used reached), was deleted. Now, I don't know what the underlying issue is, so you will have to tell me, but the symptoms are: 1. Two clients have the same groups assigned to them. 2. Group policy permitting access to exit node just, disappears -- this is important as once we set up the groups and policies we tend not to touch them. 3. One client somehow still can connect to the exit node, the other can't. I can only think of some migration problem on a DB side in the management server because this doesn't make sense. I've seen in the past how, a port number, for instance, applied to the access policy, but doesn't show in the UI. I will close this as “fixed” but how it has broken in a first place I have no clue.
Sign in to join this conversation.
No Label triage-needed
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: DYNR/netbird#10721