[Info-vax] SSH / SFTP troubles

[GerMarsh]

On Thursday, August 2, 2012 3:05:11 PM UTC+1, Steven Schweda wrote:
> > Using OpenText Secure Shell [...]
> 
> 
> 
>    Using _what_?
> 
This used to be known as Hummingbird and is a Windows 7 client. Using password authentication it works a treat!
> 
> 
> > Tried checking everything [...]
> 
> 
> 
>    Actual information might be more helpful than a vague
> 
> description.
> 
> 
Sorry about the lack of info - I meant I followed the HP SSH guide including setting up AUTHORIZATION and IDENTIFICATION files.
> 
> > [...] I did get confused as to whether I need the SSH_ADD
> 
> > to add the key though.
> 
> 
> 
>    Whether you need the _what_ to do _what_?
> 
> 
I thought that this is needed to add the private key to a repository. Maybe not.
> 
> > I tried debugging on both sides [...]
> 
> 
> 
>    Actual information might be more helpful than a vague
> 
> description.
> 
> 
I've posted some relevant info and as Hoff states, it would be useful if the debugging reported that it is checking keys etc.
> 
> > [...] discovered the logical name on the server side [...]
> 
> 
> 
>    Discovered _which_ logical name?
> 
This is the one from the SSH service command procedure: TCPIP$SSH_RUN.COM:
define/sys tcpip$ssh_server_debug y

> 
> 
> 
> 
> > [...] SFTP is utterly broken in that version. [...]
> 
> 
> 
>    Who brought SFTP into this conversation?
> 
> 
The idea is to use SFTP - setting up SSH is just the first step.
> 
> 
> 
>    I probably set up my stuff in an earlier version, and
> 
> upgraded into this one, but (omitting some blank lines):
> 
> 
> 
> alp $ ssh alp-l tcpip show version
> 
> 
> 
> @ SYS$MANAGER:ANNOUNCE.TXT      [I always enjoy that.]
> 
> 
> 
> Authentication successful.
> 
> 
> 
>   HP TCP/IP Services for OpenVMS Alpha Version V5.6 - ECO 5
> 
>   on a COMPAQ Professional Workstation XP1000 running OpenVMS V8.3
> 
> 
> 
> alp $
> 
> 
> 
Congratulations! Especially if it is only using public key authentication.
> 
> 
>    As usual, showing actual commands with their actual output
> 
> can be more helpful than vague descriptions or
> 
> interpretations.  Saying that you did everything right can be
> 
> less helpful than saying exactly what you did.  Saying that
> 
> everything looks right can be less helpful than showing what
> 
> you see.