SSH Reverse Tunneling is super useful to get remote systems connected which only have very limited internet access through mobile carriers. They usually do NAT and you have no chance to connect to these sites with a dial-in VPN or other technologies that require YOU to connect to the remote system. So we just create a reverse ssh tunnel with autossh that is kept alive by the remote system itself and we connect back to the system to the ssh tunnel. Since ssh is installed anyway, that is one of the simplest and most versatile options to connect to these systems for us.

ssh tunneling can be very useful for testing or one-shot things where you quickly need access to a service that’s not directly reachable, but I wouldn’t use it as a permanent solution for anything. You quickly run into problems like:

• TLS certificates don’t work, so you get into the habit of clicking through security warnings or turning of TLS validation altogether.
• Virtual hosts don’t work
• Port conflicts when you want to access the same type of service on different remote machines, so you have to remap them and remember things like: localhost:8080 is foo:80 and localhost:8081 is bar:80
• If it’s not your infrastructure (i.e. you are an employee in a larger company), you are probably bypassing all kinds of security rules by exposing a service and your security guys will not be too happy about it if they find out.

Thanks, bookmarking that. I always get -L and -R mixed up and have to look up examples, this one looks very handy.

Great resource, thanks for sharing

The clean new syntax for jumping is great too.

ssh -J jump.example.com target.example.com

Check out https://wiki.gentoo.org/wiki/SSH_jump_host