[isabelle-dev] isatest ssh
Gerwin.Klein at nicta.com.au
Mon Nov 26 12:11:48 CET 2012
On 26/11/2012, at 9:01 PM, Makarius <makarius at sketis.net> wrote:
> On Sun, 25 Nov 2012, Gerwin Klein wrote:
>> On 20/11/2012, at 11:23 PM, Makarius <makarius at sketis.net> wrote:
>>> There is this recurrent game to have the isatest user do many manual ssh logins to update known_hosts. Getting tired of it, I just did some reading of man ssh_config and some googling. This resulted the following ~isatest/.ssh/config:
>>> Host *
>>> #see http://linuxcommando.blogspot.fr/2008/10/how-to-disable-ssh-host-key-checking.html
>>> StrictHostKeyChecking no
>>> Maybe it helps in other situations, too. Or maybe there is an ssh expert saying that this is really really bad.
>> ssh does check these keys for a reason, it is now easy for another host to pretend to be one of the servers isatest wants to access. On the other hand, it's unclear what an attacker would gain from having isatest run a large isabelle session. There are easier ways to do that ;-)
> Do these attacks also work from the free net outside TUM?
Unclear to me. You would probably have to spoof DNS on a fairly far-reaching level to exploit this from outside TUM. My guess would be that this is a lot harder than actually getting to the server itself. But I'm not really an ssh expert.
> The reasoning (or rather hope) behind the above was that for doing real non-sense you would have to be on the local network at TUM. So it is basically a switch back towards the old-fashioned ways of rsh.
I'm fine with that part. I mainly don't want to get the warning emails for each session. Maybe we should just leave off the line with /dev/null.
> BTW, the local network is totally insecure if accessed from inside. There are many ways to do non-sense, but we better don't explain that here.
Better not ;-)
The information in this e-mail may be confidential and subject to legal professional privilege and/or copyright. National ICT Australia Limited accepts no liability for any damage caused by this email or its attachments.
More information about the isabelle-dev