Has anyone heard of an OpenSSH client being compiled with an additional command-line option for password input?
I realize there are reasons to NOT do this, and I realize you can achieve the same type of thing with keys, but I am specifically looking to pass the username & password BOTH on the command line... (IE - allowing for one-shot remote commands)
currently there is a script in place of myssh that uses Expect and Tcl to capture the output of the system-default SSH, but I would like to get rid of it and use a straight binary if possible.
Any ideas? Or should I go back to brushing up on my C/C++ and compile the option into my own version called 'myssh'?
I am, and it really is quite terrible when it comes to environmental variables like columns, colors, escape sequences, and not to mention quoting.
I even tried re-creating the same script with the Python PxSSH module (using pExpect module) and came up with ultimately the same exact hangup:
This is what I get when I try to do something like running the Dell OpenManage service startup script. (I think caused by locking or trapping) Either way, this may not be the root problem, though it is the only error displayed. The root problem is that even though the script runs, the return never happens, so Expect times out the session, calling it a failure.
You have a "system" that is comprised of N servers. This system can change in size both plus and minus (from 1 to 512 - just for kicks) individual servers. These replacements/additions all have unique IP addresses, but have a broadcast response system that can identify "all the other IPs that I need to talk to". ("I" being the server)
Given MY way, I do not need to do ANYTHING except run a bash script that uses MY SSH, logs in to every server, runs whatever I want, and pushes the output back to my terminal. This version of SSH would reside on MY machine, and would be completely compliant with their SSH daemon...
Using keys, I have to at a minimum send a file (the key) to the remote machine before communication, and that may be fine for ME to do in the lab. For someone in the field, (aka NOT me) I don't want to have to explain the details of Passwordless SSH & Keys. Especially to somebody that doesn't want/need to know.