On Sat, Apr 17, 2021 at 08:11:37PM -0400, Gene Heskett wrote:
> I just checked this machine and 4 others, all of whom are mounted to
> their own named directory as subdirs of /sshnet, which work perfectly
> without a mention of it in any machines /etc/fstab. Adding the new
> machine as the last line in this $HOME/bin/mount-machines bash script:
> ------------
> #!/bin/bash
> sshfs g...@sixty40.coyote.den:/ /sshnet/sixty40
> sshfs gene@lathe:/ /sshnet/lathe
> sshfs gene@GO704:/ /sshnet/GO704
> sshfs pi@rpi4:/ /sshnet/rpi4
> sshfs gene@TLM:/ /sshnet/TLM
> sshfs gene@3dprint:/ /sshnet/3dprint
> --------------
> Thats 5 machines that Just Work, and the last says:
>
> read: Connection reset by peer
>
> sshfs has been installed, but its not running. How do I start it?
I read this about 5 times trying to figure out what you're actually
saying. I find it very confusing.
Near as I can tell after all those attempts to read it, you have ONE
client machine and SIX server machines. You are trying to set up 6
separate sshfs mounts, all from the same client machine, one to each of
your servers.
Based on the pathname of the script that you give, you appear to be trying
to do this from an interactive shell. Not at boot time, not at login
time, not using cron, not using a systemd unit, not using /etc/rc.local,
not using a sysv-rc init.d script, not using any kind of automounting.
Now, as near as I can tell from your message, your problem comes down to
this:
From an interactive shell, I run the command
sshfs gene@3dprint:/ /sshnet/3dprint
and I get the error
read: Connection reset by peer
The first probable cause that comes to mind for me is that the sshd or
the SFTP subsystem of sshd on this "3dprint" host is not working properly.
Check the logs on that host and look for anything that might tell you
more.
Also try the typical steps that one would try when debugging an sshfs
connection:
1) Does "getent hosts 3dprint" give me the correct IP address?
2) Can I ping it?
3) Can I ssh into it? As gene?
4) Can I sftp into it? As gene?
5) If any of the above fail, is there an error message that appears in the
logs on either system at the same time?
Any additional debugging step recommendations would need to be based on
yet-to-be-revealed details. E.g. there's no point trying something like
"ssh -vvv" if you can't even ping the machine.
