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

Re: [savi] 答复: Binding establishment based on data or control packets?



yaog escribió:
> -----邮件原件-----
> 发件人: marcelo bagnulo braun [mailto:marcelo at it.uc3m.es] 
> 发送时间: 2009年1月23日 7:39
> 收件人: Christian Vogt
> 抄送: Guang Yao; SAVI Mailing List
> 主题: Re: [savi] Binding establishment based on data or control packets?
>
> Christian Vogt escribió:
>   
>> On Jan 22, 2009, marcelo bagnulo braun wrote:
>>
>>     
>>> Actually what happens when we design SAVI such that it sends RSOL
>>> packets, is that the RSOL packet acts as a synchronization point, that
>>> can be used by attacker to claim the address ownership and confuse the
>>> SAVI device. In other words, the whole point of a FCFS appraoch, is that
>>> requests arrive in the natural. If we introduce the RSOL
>>> synchronization, the order is broken and we can hardly tell who is the
>>> first one to request.
>>>       
>> Marcelo -
>>
>> The potential issue that an attacker can trick a SAVI device into
>> creating a false binding already exists in FCFS without binding
>> distribution protocol.  In that case, false bindings are limited to the
>> SAVI device to which the attacker attaches directly.  Why do you think
>> the issue becomes so much worse if a binding distribution protocol
>> enabled the attacker to create a false binding also in other SAVI
>> devices?
>>     
>
> cause the difference here is that attacker knows what IP address is a 
> valuable target (i.e. the address included in the NSOL msg)
>
> The NSOL msg acts like a call for arms for the attacker to attack that 
> address.
>
> I mean, let's focus in the IPv6 case. In IPv6, the attacker cannot 
> attack all the addresses cause 2^64 are way too many addresses. So, it 
> needs some hint about which address is a good target. Issuing the RSOL 
> acts as such.
>
> Note that there is a difference between the attacker replying to the 
> NSOL issued by the end host during the DAD and the attacker replying to 
> the NSOL issued by a SAVI device. In the case of DAD, if the attacker 
> replies, the host will not sue that address. However, when the SAVI 
> issues the NSOL message, the host is already using the address, so 
> blocking the address would imply that the host cannot use the address it 
> has configured.
>
> ~Comment Start~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> ~Guang:
> 1 We are actually blocking wrong configuration. It is reasonable.
>   
I am not sure i understand
blocking a packet sent by the rightfull owner of the address (which is
what i described above) is _not_ ok


> 2 Let pass the packets will cause more trouble.
>   

letting pass the packets sent by the rightfull owner is the right thing
to do

> 3 I think it is still synch problem first.
>   

not sure what you mean

regards, marcelo


> ~ Comment End~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
>
> Finally, suppose that the SAVI device actually sends the NSOL, then what 
> does the SAVI device does it receives two NADV  replying?
> I mean, there is no way to tell who is the rightful owner, so what 
> should the SAVI device do in that case?
>
> Regards, marcelo
>
>   
>> (I am saying "potential" issue above because the issue can be avoided if
>> the SAVI device knows that the address is already being used elsewhere.)
>>
>> - Christian
>>
>>
>>
>>     
>
>
>
>   


Note Well: Messages sent to this mailing list are the opinions of the senders and do not imply endorsement by the IETF.