IETF-SSH archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Implementation-hazards list [was Re: Fixing exchange of host keys in the SSH key exchange]



On Tue, Mar 28, 2017 at 9:04 AM, Mouse <mouse%rodents-montreal.org@localhost> wrote:
[...]
Well, in many cases.  I, for example, am not at all chary about naming
OpenSSH as the implementation whose misfeature prompted me to add
-share-number to moussh (even the moussh manpage does so)

I was curious about what that was so I looked.  Quoting moussh(1):

     There is a misfeature (I would call it a bug, except that reading the
     source makes it clear it was done deliberately) in OpenSSH's server.
     (Similar issues may exist with others, but I have no knowledge of them.)
     It gratuitously refuses to permit more than ten sessions per connection.
     This means that using moussh's connection-sharing feature to connect to
     such a server will work fine until you try to open too many remote login
     sessions, at which point you will get refusals from the remote server.
     Worst of all, OpenSSH does not provide any way for the server admin to
     raise this limit; it is hardwired into the code!

That last sentence is not accurate, OpenSSH has provided a MaxSessions config option since the 5.1 (2008): https://www.openssh.com/releasenotes.html#5.1

--
Darren Tucker (dtucker at zip.com.au)
GPG key 11EAA6FA / A86E 3E07 5B19 5880 E860  37F4 9357 ECEF 11EA A6FA (new)
    Good judgement comes with experience. Unfortunately, the experience
usually comes from bad judgement.


Home | Main Index | Thread Index | Old Index