Re: [Anima-bootstrap] bootstrap purpose

[Brian E Carpenter <[e-mail]>]

On 25/05/2016 03:02, peter van der Stok wrote:
> Hi Michael,
> 
> In earlier versions of the bootstrapping the purpose was to add an unsecured node to a secured mesh network.
> Nodes on the secured network exchanged encrypted messages.
> New nodes had to learn the keys to encrypt.
> New nodes were not allowed to communicate with other secured nodes but for starting the acquisition of the keys.
> 
> Therefore the joining node needed a neighbouring proxy, because unsecured messages were not routed over the secured network.
> I don't find the concept of secured network any more and start wondering why we need the neighbouring proxy.

The secured network is the ACP, so I wouldn't expect any of that discussion in
the bootstrap work. It should be in the description of how the ACP is formed.
(In the non-ACP case, the assumption we make is that the nodes all have appropriate
certificates for (D)TLS, and those certificates will be sufficient to keep traffic
within the AN domain.)

I though the proxy was simply the way to reach the domain registrar?

    Brian

> 
> Peter
> 
> Michael Behringer (mbehring) schreef op 2016-05-24 13:23:
>> Hi Peter,
>>
>> The primary purpose of draft-ietf-anima-bootstrapping-keyinfra is to
>> deploy a domain specific key infrastructure. You're not mentioning
>> that, I assume because it's obvious?
>>
>> What a device does with that key material is outside scope of this
>> document. Although, 3.5 and 3.6 go a little step in that direction.
>> IMO we should keep the bootstrap of key material (current objective of
>> doc) clearly separate from what to do with this key material. So we
>> should probably be clearer that 3.5 are just examples.
>>
>> To me, the objectives you mention:
>>
>>> As far as I understood there are 2 objectives;
>>> 1) a packet sent by an unauthorized node is not routed through the network
>>> 2) An unauthorized node cannot interpret a packet sent by an authorized
>>> node
>>
>> are not the main objective of this work, but of course a security
>> consideration for the bootstrap process.
>>
>> Since I'm not sure I understood you fully, please let me know whether
>> we're in line!
>> Michael
>>
>>> -----Original Message-----
>>> From: Anima-bootstrap [mailto:[e-mail]] On
>>> Behalf Of peter van der Stok
>>> Sent: 24 May 2016 09:23
>>> To: Anima-bootstrap <[e-mail]>
>>> Subject: [Anima-bootstrap] bootstrap purpose
>>>
>>> Hi all,
>>>
>>> I looked again at the keyinfra draft and did not recognize an explicit
>>> description of the purpose of securing the network with the bootstrap.
>>>
>>> As far as I understood there are 2 objectives;
>>> 1) a packet sent by an unauthorized node is not routed through the network
>>> 2) An unauthorized node cannot interpret a packet sent by an authorized
>>> node
>>>
>>> Neither does the text tell us how this is achieved once the bootstrap has
>>> successfully concluded.
>>> Do we aim at a specific protocol or do we want to leave this open?
>>>
>>> -- 
>>> Peter van der Stok
>>> vanderstok consultancy
>>>
>>> _______________________________________________
>>> Anima-bootstrap mailing list
>>> [e-mail]
>>> https://www.ietf.org/mailman/listinfo/anima-bootstrap
>>
>> _______________________________________________
>> Anima-bootstrap mailing list
>> [e-mail]
>> https://www.ietf.org/mailman/listinfo/anima-bootstrap
> 
> _______________________________________________
> Anima-bootstrap mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/anima-bootstrap
>