[olsr-dev] Selfish LQ

giuseppe de marco (spam-protected)
Thu Mar 1 09:05:27 CET 2007


Bernd Petrovitsch wrote:
> On Thu, 2007-03-01 at 09:56 +0900, giuseppe de marco wrote:
>   
>> Bernd Petrovitsch wrote:
>>     
>>> On Wed, 2007-02-28 at 17:11 +0900, giuseppe de marco wrote:
>>> [...]
>>>   
>>>       
>>>> It makes sense or I am loosing something?
>>>>     
>>>>         
>>> Apart from layer 1 and layer 2 attacks (see Sven-Olas mail), one can
>>> simply announce wrong routes (via HNA, just configure it on your
>>> interface, etc.).
>>> And all routing protocols AFAIK may suffer from routers advertising
>>> illegal/wrong routes. Usually this is not a serious problem because
>>> - on border routers (with BGPv4) one blocks neighbors advertising wrong
>>> routes.
>>> - within an AS, you have *one* organization (with interest in a working
>>> network)
>>>    and if they do not manage to get the routers working securily they
>>> (both the
>>>    admins and their bosses) are screwed anyway.
>>>       
> [...]
>   
>> What do you mean with "wrong" routes? If you mean route towards nothing, 
>>     
>
> Routes that are actually not true - either of unused IP adresses/nets or
> of used IP-adresses/nets in other parts of the network.
>
>   
>> I agree: the routers in the net does the job.
>>     
>
> And every OLSR node is a router.
>
>   
>> Actually, I am talking about good route, in the sense they are valid as 
>>     
>
> You started the first mail with
> ----  snip  ----
> Suppose I am a malicious node.
> ----  snip  ----
> So I assumed that you want to distub the rest of the network. And htat s
> much easier with other means (though very probably not so subtle as with
> "miscalculated" LQ values).
>
>   
I think it's not a disturb, rather a mis-use and then an attack. I don't 
disturb the network, i.e. I don't jam it,
 I don't do DoS. I just change a value in my OLSR module and other nodes 
have nothing to do to check
if the LQs of my declared links are true.
Regards
G

>> sequence of links towards a destination, but
>> the LQ of the link is not calculate properly.
>>     
>
> 	Bernd
>   



-- 
Giuseppe De Marco, phD
Toyota Technological Institute
468-8511 Aichi 2-12-1 Hisakata, Tenpaku-ku, 
Nagoya, Japan
Email: demarco at toyota-ti dot ac dot jp
Tel (int): +81 (052)-809-1802
Skype-Id: giuseppe_dem2





More information about the Olsr-dev mailing list