Make reverse tunnel timeout configurable

[opopops]

Describe the feature

When I configure a client to connect to the server by establishing a SOCKS5 reverse tunnel in this way:

wstunnel client --remote-to-local socks socks5://0.0.0.0:1080 wss://myRemoteHost:443

When the client disconnects, the SOCKS tunnel remains open and continues to bind to 0.0.0.0:1080on the server for 3 minutes.
It seems that this delay is hardcoded here in the code

It would be interesting to make this delay configurable via a flag for the ‘server’ subcommand.

Describe the reason for such feature

The problem is that I use an NGINX reverse proxy as a TCP load balancer to distribute the load across multiple WSTunnel servers for High Availability. As long as the server remains bound to port 1080, NGINX LB considers it ‘healthy" and this causes errors when requesting HTTP URLs through the SOCKS proxy.

[erebe]

Hello,

Thanks for reporting, I will try to add an option to make this configurable in the coming week.
I am letting you know.

Out of curiosity, in which context are you using wstunnel ?

[opopops]

Awesome! Thanks a lot for this. 🙏

About my use case, I’m using WSTunnel in a professional context to enable HTTP communication between two of apps deployed on different Kubernetes clusters. The challenge is that App 1 is on a highly secure network that App 2 cannot access. To solve this, I establish a SOCKS5 tunnel between the two clusters using WSTunnel.

[erebe]

Should be good in latest release https://github.com/erebe/wstunnel/releases/tag/v10.1.9

You have a new flag on the server --remote-to-local-server-idle-timeout 10s to specify the timeout