I'm trying to connect to machine one with ssh and then connect to another machine two with ssh, but I get this error.
ssh firstname.lastname@example.org 'ssh email@example.com' stdin: is not a tty
Unix & Linux Stack Exchange is a question and answer site for users of Linux, FreeBSD and other Un*x-like operating systems. It only takes a minute to sign up.Sign up to join this community
By default, when you run a command on the remote machine using ssh, a TTY is not allocated for the remote session. This lets you transfer binary data, etc. without having to deal with TTY quirks. This is the environment provided for the command executed on
However, when you run ssh without a remote command, it DOES allocate a TTY, because you are likely to be running a shell session. This is expected by the
ssh firstname.lastname@example.org command, but because of the previous explanation, there is no TTY available to that command.
If you want a shell on
computertwo, use this instead, which will force TTY allocation during remote execution:
ssh -t email@example.com 'ssh firstname.lastname@example.org'
This is typically appropriate when you are eventually running a shell or other interactive process at the end of the ssh chain. If you were going to transfer data, it is neither appropriate nor required to add
-t, but then every ssh command would contain a data-producing or -consuming command, like:
ssh email@example.com 'ssh firstname.lastname@example.org "cat /boot/vmlinuz"'
There's a better way to use SSH as a relay: use the
ProxyCommand option. You'll need to have a key on the client machine that lets you log in into the second computer (public key is the recommended way of using SSH in most circumstances anyway). Put this in your
~/.ssh/config and run
Host computerone HostName computerone.com User user Host computertwo HostName computertwo.com User otheruser ProxyCommand ssh computerone exec nc %h %p
nc is netcat. Any of the several versions available will do.
You can use PROXY Jump option in ssh
-J [user@]host[:port] Connect to the target host by first making a ssh connection to the jump host and then establishing a TCP forwarding to the ultimate destination from there. Multiple jump hops may be specified separated by comma characters. This is a shortcut to specify a ProxyJump configuration directive.
So if I need to connect to hostB but I have to go through hostA first to get there. Normally I would
ssh hostA [user@hostA ~]$ ssh hostB
I now do this
ssh -J hostA hostB [user@hostB ~]$
You could even
, separate multiple hosts
ssh -J hostA,hostB hostC [user@hostC]
Much simpler than trying
-t to run ssh command again.
You can override an SSH config option "RequestTTY" from the command line.
My working example
/bin/bash in SSH session after running multiple commands:
#!/bin/bash ssh -o "requestTTY=yes" User@ExampleHostName "cd /home/myPathFoo/myPathBar; /bin/bash"
And now I just run
./cd-serv-one.sh for starting a needed SSH session.
You should fix the rc files on
computerone! One of them (probably
mesg y or
mesg n without first checking whether stdin is a tty.
You should replace
mesg n by
[ -t 0 ] && mesg y.
man 1 test
0 is the file number of stdin.
ssh <host> does allocate a pseudo tty on ,
ssh <host> <cmd> (without option
-l) does not. Still, some programs (
mesg, editors, password prompts) need a terminal (or a pseudo terminal).