if A can ssh B, B can ssh C. how to setup a ssh connection from A to C, then used by rpyc.ssh_connect(ssh, port)?

[Zhi-Hua Li]

As the title states. I know the concept of ssh tunnel or port forwarding. But I have no idea how to realize this, when creating the ssh connection using Plumbum.Sshmachine().  Could anybody help? 

[Zhi-Hua Li]

Ok, let me put the question simply: Does rpyc support ssh tunnel? 

After read through the docs of rpyc, it seems to me that the core function of rpyc itself is realized by ssh tunnel. But does rpyc support that in application? 

I had setup a rpyc server on my computer (A)  and use the computation nodes (C)  inside a cluster whose  main node is (B).  Previously, C can connect directly to the rpyc server on A and everything works nicely; But I was informed that the connection from computation node to outside machine (i.e. C to A) will be forbidden by the cluster manager.  The only work around I can think of is to use some ssh tunnel mechanics, since the A<->B,  B<->C connection is always ok. But I haven't find out how to realize this in rpyc, and I don't even know rpyc support this at all.  Please help me figure to  this out. Thanks in advance.