SSH Tunnel Stopped Working

[omerXfaruq]

Acknowledgements

• [X] I have written a descriptive issue title
• [X] I have searched pyngrok's documentation for an answer
• [X] I have searched Issues to see if the bug has already been reported
• [X] I have searched Stack Overflow to ensure the issue I'm experiencing has not already been discussed

Operating System

tried on macos and ubuntu

Python Version

3.12.3

pyngrok Version

7.1.6

Describe the Bug

This week my ssh tunnel created by pyngrok stopped working. I've created a fresh virtual environment and installed the latest version of pyngrok but the problem still persists.

Steps to Reproduce

from pyngrok import ngrok ngrok.set_auth_token("TOKEN") ssh_url = ngrok.connect(22, "tcp") ssh_url

Then ssh into the tunnel from your terminal.

Expected Behavior

It should connect with ssh without any issues but it's giving an error:

$ ssh -p 12821 username@6.tcp.ngrok.io Connection closed by 3.141.177.1 port 12821

[alexdlaird]

Please rerun your script with console debug logs enabled, and provide the full output in this ticket.

Example of your script with debug logging enabled:

from pyngrok import ngrok

# Setup a logger
logger = logging.getLogger()
logger.setLevel(logging.DEBUG)
handler = logging.StreamHandler()
handler.setFormatter(logging.Formatter("%(asctime)s - %(name)s - %(levelname)s - %(message)s"))
logger.addHandler(handler)

ngrok.set_auth_token("TOKEN")
ssh_url = ngrok.connect(22, "tcp")

print("ssh_url: " + ssh_url)

[omerXfaruq]

Hi

I've checked the logs and didn't find anything wrong. Then I tried another port and tested the tcp tunnel with netcat, it worked without issues.

I'll check my ssh server side of the things as I can see that tcp tunnel works fine. Thanks.

[omerXfaruq]

Though it's still confusing because the ssh tunnel suddenly stopped working, had no changes in the system. Will troubleshoot write my findings.

[alexdlaird]

Is ufw or some other firewall intercepting the usage of ngrok? There is a nightly build that uses ubuntu-latest, which it appears GitHub runners call 22.04 (so 24.04 isn't yet tested, it seems), if you can confirm this is the same version you're on. Perhaps you can look in to how your use case differs environmentally from the nightly integration tests, here's one SSH example?

[alexdlaird]

This run is from a branch that manually points GitHub to use 24.04 runners, which also succeeded: https://github.com/alexdlaird/pyngrok/actions/runs/9664659444

[omerXfaruq]

Thanks, and yes I confirm, just tested with another ubuntu system and pyngrok works fine. It seems the problem is with my specific system.

So when I use nc through local network or ngrok tunnel, the message is delivered. But when I use pyngrok tunnel it's not. So I don't think it's a problem with the firewall. I tried creating new virtual environments as well, but didn't help. I'll try to use a different python version and see how it goes.

Also until 1-2 weeks ago I was using pyngrok to access remotely without any issues, but it suddenly started to create problems, that's why it's strange. I'll investigate more.