I was thinking of redirecting all the ssh attacks
to spamd.  spamd is a program that is used to having
bad guy attaching it, so it should not effect the 
security. Then using the max-src-conn-rate to 
block them.  

My actual problem is less with ssh then the
Microsoft vpn. I trust the people who have
ssh connections to have good passwords,
It the people with vpn connections that
I don't trust. And I of course would do
the same trick with the vpn port.

For an aside, I have wondered why the bad
guys use user names like bob, john, sally.
If I was going to be a bad guy I think
I would get a cheap mailing list and
use the email address names, as my
attacking user ids.


Matthias Kilian wrote:
> On Wed, May 31, 2006 at 03:15:34PM -0400, Peter Fraser wrote:
>> Expect I was not clear.
>>
>> Someone is attacking address 1, address 2, address 3, those
>> address are all blocked with respect to ssh. , but because he
>> is attacking those addresses, I want to stop an expected attack
>> on address 4. I never want to pass ssh on address 1, address 2
>> or address 3 ever, I want to use the information that someone
>> was trying to ssh to those address to identify person as
>> an attacker.
> 
> Oh, sorry for not reading exactly.
> 
> So your problem is that you want to get state for ssh connection
> attempts to addresses 1, 2 and 3 but at the same time want to block
> those connections. This isn't possible (no connection - no state).
> 
> (QUICK HACK ALERT)
> 
> But it may be possible to redirect those connections to some unused
> port on localhost (i.e. the firewall) let something listen on this
> port, accept everything but immediately closing the connection.
> Then use a simple pass rule with overload and max-src-conn options
> to add offending addresses to your table.
> 
> Ciao,
>       Kili
> 
> ps: I didn't test the above, so if it's complete nonsense, feel
> free to flame me.

Reply via email to