What you're referring is "SSH REMOTE FORWARDING", and is properly explained in the "man ssh", regarding the "-R" option.
> man ssh
Specifies that the given port on the remote (server) host is to
be forwarded to the given host and port on the local side.
This works by allocating a socket to listen to port on the remote
side, and whenever a connection is made to this port, the
connection is forwarded over the secure channel, and a connection is
made to host port hostport from the local machine.
In your context, where:
- a Linux box A (LINUX_BOX_A) inside a LAN behind a firewall.
- a Linux server B (SERVER_B) with a fixed IP that is accessible from the internet
SSH remote forwarding can be used to reach LINUX_BOX_A from SERVER_B. The only condition is: LINUX_BOX_A MUST be able to connect via SSH to SERVER_B.
To achieve this goal you need:
- on LINUX_BOX_A:
LINUX_BOX_A:~ $ ssh -R 2222:localhost:22 user@SERVER_B
this will open an ssh connection from LINUX_BOX_A to SERVER_B that will be used for the remote, incoming, connection.
After above ssh connection is established, you can:
- on SERVER_B:
SERVER_B:~ $ ssh -p 2222 user@localhost
such ssh-connection, launched on SERVER_B, will be directed to the 2222 port listening on localhost that... is binded to the previous ssh connection. So this will be an "ssh connection within another ssh connection".
Some additional notes:
consider that if the first ssh-connection will timeout and/or fall down for whatever reason (including: killed by local firewall, due to inactivity), you'll be unable to remote-forward/remotely_connect;
as it's important to leave the first ssh connection active for really long time, you might find useful to launch such ssh within a "screen" session
A final note: Obviously, all of the above has some (potentially serious) security implications that are out of scope of this answer.