Hi Sue,

On 21/04/16 19:59, Susan Hares wrote:
> Stephen: 
> 
> I have added another section in the security consideration section: 
> 
>           4.4.  I2RS in Personal Devices      
>               
>          If an I2RS agent or I2RS client is tightly correlated with a person  
>          (such as if an I2RS agent is running on someone's phone to control   
>          tethering) then this usage can raise privacy issues, over and above  
>          the security issues normally need to be handled in I2RS.  One 
> example        
>          of an I2RS interaction that could raise privacy issues, is if the    
>          I2RS interaction enabled easier location tracking of a person's      
>          phone.  The I2RS protocol and data models should consider if privacy 
>          issues can arise when clients or agents are used for such use-cases. 
> 
> Does this address your issue with privacy issues? 

Well, given that those are almost exactly the words I suggested
below, sure, they're good enough:-)

Cheers,
S.

> 
> Sue Hares 
> 
> -----Original Message-----
> From: Stephen Farrell [mailto:[email protected]] 
> Sent: Thursday, March 17, 2016 9:28 AM
> To: Joel Halpern; Susan Hares; 'The IESG'
> Cc: [email protected]; [email protected]; [email protected]; 
> [email protected]
> Subject: Re: Stephen Farrell's No Objection on 
> draft-ietf-i2rs-architecture-13: (with COMMENT)
> 
> 
> 
> On 17/03/16 13:25, Joel Halpern wrote:
>> Can you suggest wording to add to the architecture document to reflect this 
>> consideration?
> 
> Maybe something along the lines of:
> 
> "If an i2rs agent or client is such that it is likely
> tightly correlated with a person (say if an agent is
> running on someone's phone to control tethering) then
> that can raise privacy issues, over and above.the
> security and privacy issues that normally need to be
> handled in i2rs. For example, if an i2rs interaction
> enabled easier location tracking in the above example.
> i2rs protocols should consider if such privacy issues
> can arise when clients or agents are used for such
> use-cases."
> 
> Cheers
> S.
> 
> 
>>
>> Yours,
>> Joel
>>
>> -----Original Message-----
>> From: Stephen Farrell [mailto:[email protected]]
>> Sent: Thursday, March 17, 2016 2:23 PM
>> To: Joel Halpern; Susan Hares; 'The IESG'
>> Cc: [email protected]; [email protected]; [email protected]; 
>> [email protected]
>> Subject: Re: Stephen Farrell's No Objection on 
>> draft-ietf-i2rs-architecture-13: (with COMMENT)
>>
>>
>>
>> On 17/03/16 13:15, Joel Halpern wrote:
>>> I would hope that I2RS could be used for that (applying policy to home
>>> devices) use case.
>>
>> Ah. Good to know.
>>
>>>
>>> But I am not at all clear how I2RS could protect the IP address of the 
>>> router
>>> containing the communicating I2RS agent.  We have to have an available IP
>>> address for IP Routing.
>>
>> I didn't say it needed protecting (as in encrypting) necessarily,
>> but that it could be more sensitive.
>>
>>>
>>> I am also not clear why this IP address is particularly more sensitive than 
>>> an
>>> enterprise device IP address, or a router inside an ISP.
>>
>> In general, if an identifier is also something one can correlate
>> with a person, or with a person's movements or presence, then it
>> is more privacy sensitive. If you can tell I'm at home because of
>> an i2rs event say.
>>
>> For a router on the 4th floor of an office building, those are
>> less likely interesting issues.
>>
>> In the home case, one needs to think more about such stuff than
>> in the office case basically.
>>
>> Whether/how that impacts on protocol design is hard to say. But
>> it's good to know that it's something that i2rs needs to consider.
>>
>> Cheers,
>> S.
>>
>>
>>>
>>> Yours,
>>> Joel
>>>
>>> -----Original Message-----
>>> From: Stephen Farrell [mailto:[email protected]]
>>> Sent: Thursday, March 17, 2016 2:11 PM
>>> To: Susan Hares; 'The IESG'
>>> Cc: [email protected]; [email protected];
>>> [email protected]; [email protected]
>>> Subject: Re: Stephen Farrell's No Objection on
>>> draft-ietf-i2rs-architecture-13: (with COMMENT)
>>>
>>>
>>> Hiya,
>>>
>>> Just on that one point (the rest seems fine):
>>>
>>> On 17/03/16 13:00, Susan Hares wrote:
>>>>>> - If i2rs were used to control home networks, then that would
>>>>>> raise more privacy issues, e.g. the agent's IP address can be
>>>>>> privacy sensitive. Would it be useful to rule that out of
>>>>> scope? E.g. to say that i2rs SHOULD NOT be used where the
>>>>> agent/router in question
>>>>>> is specific to one person or home?
>>>
>>>> Sue:  I'm really not sure what you are getting at.  Data in routers
>>>> is privacy sensitive. Data between I2RS Agent and I2RS client will be
>>>> encrypted except in very, very rare circumstances where is defined to
>>>> be public data in the data model. SECDIR, OPSDIR, RTGWG,
>>>> Transport-directorate will be asked to review any IETF data model
>>>> that claims this is the case to validate it is appropriate.   So... I
>>>> think we are going beyond what people use for home networks.
>>>
>>> Let's assume all client/agent stuff is wonderfully protected
>>> e.g. via TLS.
>>>
>>> Normally, the fact that a client at IP1 is managing an agent at
>>> IP2, which is still visible despite the TLS, is not much of a
>>> deal. Nor is it a deal when that happens, e.g. in reaction to
>>> some other event, perhaps even one triggered by an attacker.
>>>
>>> But if the agent is my home g/w, then the sensitivity level goes
>>> up I think, or at least it can. The reason is that the agent's
>>> address (IP2) is tied to me. If the agent was on my phone (e.g.
>>> for tethering) then it'd be even more of a deal perhaps, as I
>>> carry it with me.
>>>
>>> If i2rs just isn't intended for such use-cases, it may be worth
>>> saying that was all I meant.
>>>
>>> Cheers,
>>> S.
>>>
>>>
>>>>
>>>
>>
> 
> 

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

_______________________________________________
i2rs mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/i2rs

Reply via email to