in reply to Re: GRID::Machine with authentication
in thread GRID::Machine with authentication
I still don't see the problem. This is most likely because you haven't described the whole situation. I think a current solution would still be to give everybody their own, passwordless key and make them use their personal key when they launch the script.
I don't see why your construct of multiple SSH/telnet hops is necessary, but it can easily be eliminated by launching netcat as the remote shell to do the direct ssh tunneling.
|
|---|