Windows client not connect #2298

Open
opened 2025-11-20 07:07:16 -05:00 by saavagebueno · 32 comments
Owner

Originally created by @sgm456 on GitHub (Sep 22, 2025).

Originally assigned to: @pappz on GitHub.

The Windows client doesn't connect to the Netbird server on versions 0.58 and 0.58.1.
The latest working version is 0.57.1.

A clear and concise description of what the problem is.

To Reproduce

Steps to reproduce the behavior:

  1. Install agent on windows
  2. See error
Originally created by @sgm456 on GitHub (Sep 22, 2025). Originally assigned to: @pappz on GitHub. The Windows client doesn't connect to the Netbird server on versions 0.58 and 0.58.1. The latest working version is 0.57.1. A clear and concise description of what the problem is. **To Reproduce** Steps to reproduce the behavior: 1. Install agent on windows 4. See error
saavagebueno added the triage-needed label 2025-11-20 07:07:16 -05:00
Author
Owner

@sunstarjeff commented on GitHub (Sep 22, 2025):

I came to the Issues page today to see if anyone had reported this issue, after experiencing connectivity issues on several Windows clients.

I'm reverting Windows clients back to 0.57.1 until this issue can be resolved.

@sunstarjeff commented on GitHub (Sep 22, 2025): I came to the Issues page today to see if anyone had reported this issue, after experiencing connectivity issues on several Windows clients. I'm reverting Windows clients back to 0.57.1 until this issue can be resolved.
Author
Owner

@mlsmaycon commented on GitHub (Sep 22, 2025):

@sgm456 @sunstarjeff could you please share the upload key for the following command in recent versions:

netbird debug bundle --upload-bundle --anonymize
@mlsmaycon commented on GitHub (Sep 22, 2025): @sgm456 @sunstarjeff could you please share the upload key for the following command in recent versions: ```shell netbird debug bundle --upload-bundle --anonymize ```
Author
Owner

@sunstarjeff commented on GitHub (Sep 22, 2025):

Sure, is this what you're needing?

C:\WINDOWS\SystemTemp\netbird.debug.1785713459.zip
Upload file key:
f79e391890ab27fb37c88b3b4be7011e22aa2e5ca6f38ffa9c4481884941f726/fe13d424-65ed-4442-81f7-cbfa604d1254

@sunstarjeff commented on GitHub (Sep 22, 2025): Sure, is this what you're needing? C:\WINDOWS\SystemTemp\netbird.debug.1785713459.zip Upload file key: f79e391890ab27fb37c88b3b4be7011e22aa2e5ca6f38ffa9c4481884941f726/fe13d424-65ed-4442-81f7-cbfa604d1254
Author
Owner

@sgm456 commented on GitHub (Sep 23, 2025):

C:\tmp>netbird debug bundle --upload-bundle --anonymize Local file: C:\WINDOWS\SystemTemp\netbird.debug.1101652429.zip Upload file key: f79e391890ab27fb37c88b3b4be7011e22aa2e5ca6f38ffa9c4481884941f726/42c5f731-21bb-431c-a2fa-21c56434ee88

@sgm456 commented on GitHub (Sep 23, 2025): C:\tmp>netbird debug bundle --upload-bundle --anonymize Local file: C:\WINDOWS\SystemTemp\netbird.debug.1101652429.zip Upload file key: f79e391890ab27fb37c88b3b4be7011e22aa2e5ca6f38ffa9c4481884941f726/42c5f731-21bb-431c-a2fa-21c56434ee88
Author
Owner

@lixmal commented on GitHub (Sep 23, 2025):

Can you please run netbird debug for 1m -UA instead? Then share the upload file key again.
@sgm456 @sunstarjeff

Can you also describe what the issue is exactly? "See error" is not helpful

@lixmal commented on GitHub (Sep 23, 2025): Can you please run `netbird debug for 1m -UA` instead? Then share the upload file key again. @sgm456 @sunstarjeff Can you also describe what the issue is exactly? "See error" is not helpful
Author
Owner

@mrwsl commented on GitHub (Sep 23, 2025):

Our windows clients experience the same on version 0.58.0, while linux clients are not affected. Downgrading the windows clients to 0.57.1 helps. Here is the debug upload bundle:

Upload file key:
a43af7f29c88cc80d10dfbf4f3396dd4829a71a3edc2a76a38c81b747a819736/26b072dc-a7cd-4b1c-8fce-6e3a5b5b5a1e

EDIT: The "error" is that no device is connected to the client. They are showing 0/x connected.

@mrwsl commented on GitHub (Sep 23, 2025): Our windows clients experience the same on version 0.58.0, while linux clients are not affected. Downgrading the windows clients to 0.57.1 helps. Here is the debug upload bundle: ``` Upload file key: a43af7f29c88cc80d10dfbf4f3396dd4829a71a3edc2a76a38c81b747a819736/26b072dc-a7cd-4b1c-8fce-6e3a5b5b5a1e ``` EDIT: The "error" is that no device is connected to the client. They are showing `0/x connected`.
Author
Owner

@lixmal commented on GitHub (Sep 23, 2025):

Thanks @mrwsl

For the time being you can downgrade like you did or set this env var:

netbird service reconfigure --service-env NB_USE_LEGACY_ROUTING=true

@lixmal commented on GitHub (Sep 23, 2025): Thanks @mrwsl For the time being you can downgrade like you did or set this env var: `netbird service reconfigure --service-env NB_USE_LEGACY_ROUTING=true`
Author
Owner

@sgm456 commented on GitHub (Sep 23, 2025):

C:\tmp>netbird service reconfigure --service-env NB_USE_LEGACY_ROUTING=true
Stopping NetBird service...
Removing existing service configuration...
Installing service with new configuration...
Starting NetBird service...
NetBird service has been reconfigured and started

C:\tmp>netbird debug for 1m -UA
Error: failed to set log level to TRACE: engine not initialized

C:\tmp>netbird debug for 1m -UA
netbird down
Error: failed to up: up already in progress: current status LoginFailed

After reverting to 0.57.1, I connected successfully.

@sgm456 commented on GitHub (Sep 23, 2025): C:\tmp>netbird service reconfigure --service-env NB_USE_LEGACY_ROUTING=true Stopping NetBird service... Removing existing service configuration... Installing service with new configuration... Starting NetBird service... NetBird service has been reconfigured and started C:\tmp>netbird debug for 1m -UA Error: failed to set log level to TRACE: engine not initialized C:\tmp>netbird debug for 1m -UA netbird down Error: failed to up: up already in progress: current status LoginFailed After reverting to 0.57.1, I connected successfully.
Author
Owner

@mlsmaycon commented on GitHub (Sep 23, 2025):

@sgm456 can you confirm that the issue happens after a few seconds of running the reconfigure command? I mean, if you wait, 10-30 seconds before running netbird debug for 1m -U -A does it work?

@mlsmaycon commented on GitHub (Sep 23, 2025): @sgm456 can you confirm that the issue happens after a few seconds of running the reconfigure command? I mean, if you wait, 10-30 seconds before running `netbird debug for 1m -U -A` does it work?
Author
Owner

@sgm456 commented on GitHub (Sep 23, 2025):

@sgm456 can you confirm that the issue happens after a few seconds of running the reconfigure command? I mean, if you wait, 10-30 seconds before running netbird debug for 1m -U -A does it work?

After 1 min connected. 0.58.1

@sgm456 commented on GitHub (Sep 23, 2025): > [@sgm456](https://github.com/sgm456) can you confirm that the issue happens after a few seconds of running the reconfigure command? I mean, if you wait, 10-30 seconds before running `netbird debug for 1m -U -A` does it work? After 1 min connected. 0.58.1
Author
Owner

@mlsmaycon commented on GitHub (Sep 23, 2025):

@sgm456 thanks for the update. Can you generate the bundle with?

netbird debug for 2m -U -A
@mlsmaycon commented on GitHub (Sep 23, 2025): @sgm456 thanks for the update. Can you generate the bundle with? ```shell netbird debug for 2m -U -A ```
Author
Owner

@davidchi2020 commented on GitHub (Sep 23, 2025):

I also encountered a very similar issue and noticed a possible pattern: After upgrading the NetBird client on one of my Windows PCs (PC1) from version 0.57.1 to 0.58.0 or 0.58.1, it lost connectivity with some other NetBird peers. These unreachable peers were running either Windows or Linux, and the connectivity issue didn’t seem related to their NetBird client versions. However, I found that if the unreachable peers actively pinged PC1’s NetBird domain name, the connection between them would be restored. Not sure if this observation could provide useful clues for troubleshooting.

@davidchi2020 commented on GitHub (Sep 23, 2025): I also encountered a very similar issue and noticed a possible pattern: After upgrading the NetBird client on one of my Windows PCs (PC1) from version 0.57.1 to 0.58.0 or 0.58.1, it lost connectivity with some other NetBird peers. These unreachable peers were running either Windows or Linux, and the connectivity issue didn’t seem related to their NetBird client versions. However, I found that if the unreachable peers actively pinged PC1’s NetBird domain name, the connection between them would be restored. Not sure if this observation could provide useful clues for troubleshooting.
Author
Owner

@davidchi2020 commented on GitHub (Sep 23, 2025):

I also encountered a very similar issue and noticed a possible pattern: After upgrading the NetBird client on one of my Windows PCs (PC1) from version 0.57.1 to 0.58.0 or 0.58.1, it lost connectivity with some other NetBird peers. These unreachable peers were running either Windows or Linux, and the connectivity issue didn’t seem related to their NetBird client versions. However, I found that if the unreachable peers actively pinged PC1’s NetBird domain name, the connection between them would be restored. Not sure if this observation could provide useful clues for troubleshooting.

On two Windows 10 PCs, after upgrading NetBird from version 0.57.1 to 0.58.1, they were unable to establish a connection. Pinging each other did not help either. It seems the pattern I mentioned earlier was inaccurate—this is quite puzzling.

@davidchi2020 commented on GitHub (Sep 23, 2025): > I also encountered a very similar issue and noticed a possible pattern: After upgrading the NetBird client on one of my Windows PCs (PC1) from version 0.57.1 to 0.58.0 or 0.58.1, it lost connectivity with some other NetBird peers. These unreachable peers were running either Windows or Linux, and the connectivity issue didn’t seem related to their NetBird client versions. However, I found that if the unreachable peers actively pinged PC1’s NetBird domain name, the connection between them would be restored. Not sure if this observation could provide useful clues for troubleshooting. On two Windows 10 PCs, after upgrading NetBird from version 0.57.1 to 0.58.1, they were unable to establish a connection. Pinging each other did not help either. It seems the pattern I mentioned earlier was inaccurate—this is quite puzzling.
Author
Owner

@sgm456 commented on GitHub (Sep 23, 2025):

C:\tmp>netbird debug for 2m -U -A netbird down netbird up Remaining time: 00:00:01 Duration completed Creating debug bundle... Local file: C:\WINDOWS\SystemTemp\netbird.debug.1260464448.zip Upload file key: f79e391890ab27fb37c88b3b4be7011e22aa2e5ca6f38ffa9c4481884941f726/445cb134-ba46-4b9e-82a1-3586402b0df1

@sgm456 commented on GitHub (Sep 23, 2025): C:\tmp>netbird debug for 2m -U -A netbird down netbird up Remaining time: 00:00:01 Duration completed Creating debug bundle... Local file: C:\WINDOWS\SystemTemp\netbird.debug.1260464448.zip Upload file key: f79e391890ab27fb37c88b3b4be7011e22aa2e5ca6f38ffa9c4481884941f726/445cb134-ba46-4b9e-82a1-3586402b0df1
Author
Owner

@lixmal commented on GitHub (Sep 24, 2025):

It's not clear yet what's wrong, according to the log the correct outbound interface is selected etc

@sgm456 you seem to have the legacy routing env var enabled
@mrwsl all peers are in Idle state so we don't get any valuable info

Please follow these steps and generate another bundle:

  1. If activated, deactivate lazy connections. Or if that's not desirable, then ping one of the affected peers while the netbird debug runs to activate it
  2. Ensure legacy routing is off if previously enabled: netbird service reconfigure --service-env NB_USE_LEGACY_ROUTING=false
  3. Run netbird client v0.58.1
  4. If possible, capture traffic on the wt0 interface via Wireshark or pktmon before/during the test
  5. Run netbird debug for 2m -AU
  6. Share the upload key, the pcap file from the capture and the Windows version with patch level
@lixmal commented on GitHub (Sep 24, 2025): It's not clear yet what's wrong, according to the log the correct outbound interface is selected etc @sgm456 you seem to have the legacy routing env var enabled @mrwsl all peers are in `Idle` state so we don't get any valuable info Please follow these steps and generate another bundle: 1. If activated, deactivate lazy connections. Or if that's not desirable, then ping one of the affected peers while the `netbird debug` runs to activate it 2. Ensure legacy routing is off if previously enabled: `netbird service reconfigure --service-env NB_USE_LEGACY_ROUTING=false` 3. Run netbird client `v0.58.1` 4. If possible, capture traffic on the `wt0` interface via Wireshark or [pktmon](https://learn.microsoft.com/en-us/windows-server/networking/technologies/pktmon/pktmon) before/during the test 5. Run `netbird debug for 2m -AU` 6. Share the upload key, the `pcap` file from the capture and the Windows version with patch level
Author
Owner

@mrwsl commented on GitHub (Sep 25, 2025):

Upload file key:
a43af7f29c88cc80d10dfbf4f3396dd4829a71a3edc2a76a38c81b747a819736/2a1bd141-29e2-49ac-a486-991a057273a2

I pinged a peer while capturing traffic on wt0 (feel free to delete the Wireshark capture attachment, once you have it).

@mrwsl commented on GitHub (Sep 25, 2025): ``` Upload file key: a43af7f29c88cc80d10dfbf4f3396dd4829a71a3edc2a76a38c81b747a819736/2a1bd141-29e2-49ac-a486-991a057273a2 ``` I pinged a peer while capturing traffic on `wt0` (feel free to delete the Wireshark capture attachment, once you have it).
Author
Owner

@sgm456 commented on GitHub (Sep 25, 2025):

C:>netbird service reconfigure --service-env NB_USE_LEGACY_ROUTING=false
Stopping NetBird service...
Removing existing service configuration...
Installing service with new configuration...
Starting NetBird service...
NetBird service has been reconfigured and started

C:>netbird debug for 2m -AU
Error: failed to set log level to TRACE: engine not initialized

C:>netbird up
Error: daemon up failed: login backoff cycle failed: rpc error: code = Unknown desc = getting device authorization flow info failed with error: context deadline exceeded

C:>netbird debug for 2m -AU
Error: failed to up: up already in progress: current status LoginFailed

@sgm456 commented on GitHub (Sep 25, 2025): C:\>netbird service reconfigure --service-env NB_USE_LEGACY_ROUTING=false Stopping NetBird service... Removing existing service configuration... Installing service with new configuration... Starting NetBird service... NetBird service has been reconfigured and started C:\>netbird debug for 2m -AU Error: failed to set log level to TRACE: engine not initialized C:\>netbird up Error: daemon up failed: login backoff cycle failed: rpc error: code = Unknown desc = getting device authorization flow info failed with error: context deadline exceeded C:\>netbird debug for 2m -AU Error: failed to up: up already in progress: current status LoginFailed
Author
Owner

@davidchi2020 commented on GitHub (Sep 26, 2025):

The issue only occurs with versions starting from 0.58.0 up to 0.58.2. I kept pinging a peer in the command line window. After upgrading from 0.57.1 to 0.58.0 (or 0.58.1, 0.58.2), the ping stopped receiving responses. Then, after reinstalling version 0.57.1, the ping started receiving responses again.

@davidchi2020 commented on GitHub (Sep 26, 2025): The issue only occurs with versions starting from 0.58.0 up to 0.58.2. I kept pinging a peer in the command line window. After upgrading from 0.57.1 to 0.58.0 (or 0.58.1, 0.58.2), the ping stopped receiving responses. Then, after reinstalling version 0.57.1, the ping started receiving responses again.
Author
Owner

@sgm456 commented on GitHub (Oct 2, 2025):

0.59, 0.59.1 the problem remains

@sgm456 commented on GitHub (Oct 2, 2025): 0.59, 0.59.1 the problem remains
Author
Owner

@lbios1 commented on GitHub (Oct 3, 2025):

0.59.2 doesn't work either

@lbios1 commented on GitHub (Oct 3, 2025): 0.59.2 doesn't work either
Author
Owner

@sgm456 commented on GitHub (Oct 7, 2025):

0.59.3 doesn't work either too

@sgm456 commented on GitHub (Oct 7, 2025): 0.59.3 doesn't work either too
Author
Owner

@matkoeout commented on GitHub (Oct 7, 2025):

I have the same problem with many clients. Since version 58 all clients can´t conntect to peers.

@matkoeout commented on GitHub (Oct 7, 2025): I have the same problem with many clients. Since version 58 all clients can´t conntect to peers.
Author
Owner

@davidchi2020 commented on GitHub (Oct 9, 2025):

After upgrading to version 0.59.4, the issue still persists.

@davidchi2020 commented on GitHub (Oct 9, 2025): After upgrading to version 0.59.4, the issue still persists.
Author
Owner

@davidchi2020 commented on GitHub (Oct 9, 2025):

Image The client logs showed that all peers remained Idle, and connections could not be activated by pinging them. The issue was resolved by toggling the "lazy connections" feature off globally in the dashboard. This confirms the problem is a bug in lazy state management, present from version 0.58.0 through to the current 0.59.5.
@davidchi2020 commented on GitHub (Oct 9, 2025): <img width="1475" height="438" alt="Image" src="https://github.com/user-attachments/assets/05208b6f-f1e1-4706-86c3-47e57bc2e899" /> The client logs showed that all peers remained Idle, and connections could not be activated by pinging them. The issue was resolved by toggling the "lazy connections" feature off globally in the dashboard. This confirms the problem is a bug in lazy state management, present from version 0.58.0 through to the current 0.59.5.
Author
Owner

@sevensolutions commented on GitHub (Oct 10, 2025):

@davidchi2020 can confirm.
As soon as i disable lazy connections in the dashbaord, it starts to work immediately.

@sevensolutions commented on GitHub (Oct 10, 2025): @davidchi2020 can confirm. As soon as i disable lazy connections in the dashbaord, it starts to work immediately.
Author
Owner

@sgm456 commented on GitHub (Oct 12, 2025):

After upgrading to version 0.59.5, the issue still persists.

@sgm456 commented on GitHub (Oct 12, 2025): After upgrading to version 0.59.5, the issue still persists.
Author
Owner

@zanderson-aim commented on GitHub (Oct 13, 2025):

Confirmed as well, disable Lazy Connections Globally and we are good

@zanderson-aim commented on GitHub (Oct 13, 2025): Confirmed as well, disable Lazy Connections Globally and we are good
Author
Owner

@lixmal commented on GitHub (Oct 17, 2025):

Hey folks, this should be fixed with https://github.com/netbirdio/netbird/releases/tag/v0.59.7

@lixmal commented on GitHub (Oct 17, 2025): Hey folks, this should be fixed with https://github.com/netbirdio/netbird/releases/tag/v0.59.7
Author
Owner

@zanderson-aim commented on GitHub (Oct 17, 2025):

Looks good here

@zanderson-aim commented on GitHub (Oct 17, 2025): Looks good here
Author
Owner

@sgm456 commented on GitHub (Oct 18, 2025):

I figured out the problem.
The problem:
Netbird versions greater than 0.57.1 fail to connect to the server with a "login failed" error.
Cause: Dr.Web Security Space 12 (for workstations)
Disabling all protection components doesn't produce a positive result.
However, if you completely uninstall the antivirus or add Netbird to the exceptions, the connection is established.
The same connection issue occurs with some websites when Dr.Web Security Space is enabled, for example https://pixeldrain.com/ (ERR_SSL_PROTOCOL_ERROR).
It's unclear why the antivirus reacted this way to new versions of Netbird. Something in the TLS/SSL settings in the Netbird client has changed.
It's also unclear why the latest versions of Netbird, when running:
netbird service reconfigure --service-env NB_USE_LEGACY_ROUTING=true, still connect without exceptions in the antivirus.
In general, the cause on the workstation has been determined.

@sgm456 commented on GitHub (Oct 18, 2025): I figured out the problem. The problem: Netbird versions greater than 0.57.1 fail to connect to the server with a "login failed" error. Cause: Dr.Web Security Space 12 (for workstations) Disabling all protection components doesn't produce a positive result. However, if you completely uninstall the antivirus or add Netbird to the exceptions, the connection is established. The same connection issue occurs with some websites when Dr.Web Security Space is enabled, for example https://pixeldrain.com/ (ERR_SSL_PROTOCOL_ERROR). It's unclear why the antivirus reacted this way to new versions of Netbird. Something in the TLS/SSL settings in the Netbird client has changed. It's also unclear why the latest versions of Netbird, when running: netbird service reconfigure --service-env NB_USE_LEGACY_ROUTING=true, still connect without exceptions in the antivirus. In general, the cause on the workstation has been determined.
Author
Owner

@Kimmax commented on GitHub (Oct 18, 2025):

This is not the problem, login is successful and no "Dr. Web Security" installed.
@lixmal Could you link the commit that fixed this? I'd like to understand what was going on :)

@Kimmax commented on GitHub (Oct 18, 2025): This is not the problem, login is successful and no "Dr. Web Security" installed. @lixmal Could you link the commit that fixed this? I'd like to understand what was going on :)
Author
Owner

@lixmal commented on GitHub (Oct 18, 2025):

@Kimmax https://github.com/netbirdio/netbird/pull/4646

@lixmal commented on GitHub (Oct 18, 2025): @Kimmax https://github.com/netbirdio/netbird/pull/4646
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: SVI/netbird#2298