Re: [lisp] Kathleen Moriarty's No Objection on draft-ietf-lisp-impact-04: (with COMMENT)

[Luigi Iannone <[e-mail]>]

Hi Kathleen,

> On 21 Oct 2015, at 18:52, Kathleen Moriarty <[e-mail]> wrote:
> 
> 

[snip]

> 
> 
> ----------------------------------------------------------------------
> COMMENT:
> ----------------------------------------------------------------------
> 
> Hello,
> 
> There was no follow up or changes (it seems) as a result of the SecDir
> review.  It would be helpful to address the questions on the aim of this
> draft and how it applies to security for the user and impact of LISP.
> https://www.ietf.org/mail-archive/web/secdir/current/msg06103.html
> 
> 

There was actually a follow up  (see below) or ami I missing something?

Let me know.

ciao

L.


%—— Last reply to Hilarie on 20th October———————%
Hi Hilarie,

Thanks again for your reply.
please find our comments inline.

ciao

Luigi


> On 19 Oct 2015, at 21:02, Hilarie Orman <[e-mail]> wrote:
> 
> [NB: this is in re draft-ietf-lisp-impact-04]
> 
> A few comments and suggestions:
> 
>    Unless gleaning features (actually deprecated in
>    RFC 6830 [RFC6830]) are used, 
> 
> I don't see that gleaning is deprecated.  In any event, how does gleaning
> undermine security?

This is actually discussed in sections 6 and 12 of RFC6830 and analysed in Section 3.1 of draft-ietf-lisp-threats.

> 
>                                   the  LISP data-plane shows the 
>    same level of security as other IP-over-IP technologies.
>    From a security perspective, the control-plane remains the 
>    critical part of the LISP architecture.
> 
>    To maximally mitigate the threats on the mapping
> 
> I doubt authentication is "maximal" mitigation.  It just mitigates.

Agreed. The sentence will be simplified as just “To mitigate the threats…."

> 
>    system, authentication must be used, whenever possible, for all 
> 
> When would it be impossible to use authentication?
> 

The idea was to hint at deployments in ressource constrained environments.
It might in fact be misleading. The whole sentence can be reworded as follows:

	To mitigate the threats on the mapping system, authentication 
	should be used for all control plane messages.


>    control plane messages.
> 
>    Current specification already offer security mechanisms 
>    ([RFC6833],  [I-D.ietf-lisp-sec]) able to strongly reduce threats 
>    in non-trustable environments such as the Internet.  
> 
> "The currenet specification defines security mechanisms which can
> reduce threats in open network environments”

Just to keep the references, the sentence can be:

	The current specification ([RFC6833],  [I-D.ietf-lisp-sec]) defines security 
	mechanisms which can reduce threats in open network environments. 


> ?
> 

>    Actually, LISP specifications define a generic authentication data field 
>    control plane messages [RFC6830] allowing to propose a general
>    authentication mechanisms for the LISP control-plane while staying
>    backward compatible. 
> 
> "The LISP specification defines a generic authentication data field 
>    for control plane messages [RFC6830] which could be used for a general
>    authentication mechanisms for the LISP control-plane while staying
>    backward compatible. "  ??
> 

Reads much better, thanks.

Luigi

> Hilarie
> 
>> Subject: Re: review of draft-saucez-lisp-impact-04.txt
>> From: Luigi Iannone <[e-mail]>
>> Date: Sat, 17 Oct 2015 21:49:24 +0200
>> Cc: Damien Saucez <[e-mail]>,
>> 	   [e-mail], [e-mail],
>> 	   The IESG <[e-mail]>
> 
>> Hi Hilarie,
> 
>> In the current format the security section just states that actually 
>> security is out of the scope of the document.
>> This was actually an outcome of the WG discussion, were it was
>> decided to clearly separate security and impact.
> 
> 
>> Yet, it is true that the security section is poor, while 
>> security analysis is out of the scope of the document, it does not 
>> mean that we cannot mention the major security points 
>> thoroughly analysed in the threats document.
> 
> 
>> Hence we propose to modify the security section as follows:
> 
>> Old Version:
> 
>> 	   Security and threats analysis of the LISP protocol is out of the
>> 	   scope of the present document.  A thorough analysis of LISP security
>> 	   threats is detailed in [I-D.ietf-lisp-threats].
> 
> 
>> NEW Version:
> 
>> 	   A thorough security and threats analysis of the LISP protocol
>> 	   is carried out in details in [I-D.ietf-lisp-threats]. 
>> 	   Like for other Internet technologies, also for LISP most of 
>> 	   threats can be mitigated using Best Current Practice, meaning 
>> 	   with careful deployment an configuration (e.g., filter) and also 
>> 	   by activating only features that are really necessary in the 
>> 	 deployment and verifying all the information obtained from third 
>> 	   parties. Unless gleaning features (actually deprecated in
>> 	   RFC 6830 [RFC6830]) are used, the  LISP data-plane shows the 
>> 	   same level of security as other IP-over-IP technologies.
>> 	   From a security perspective, the control-plane remains the 
>> 	   critical part of the LISP architecture.
>> 	   To maximally mitigate the threats on the mapping
>> 	   system, authentication must be used, whenever possible, for all 
>> 	   control plane messages.
>> 	   Current specification already offer security mechanisms 
>> 	   ([RFC6833],  [I-D.ietf-lisp-sec]) able to strongly reduce threats 
>> 	   in non-trustable environments such as the Internet.  
>> 	   Actually, LISP specifications define a generic authentication data field 
>> 	   control plane messages [RFC6830] allowing to propose a general
>> 	   authentication mechanisms for the LISP control-plane while staying
>> 	   backward compatible. 
> 
> 
>> We hope this delivers the information you were looking for.
> 
>> ciao
> 
>> Luigi
> 
> 
>>> On 13 Oct 2015, at 19:28, Hilarie Orman <[e-mail]> wrote:
>>> 
>>> Thanks for pointing out my mistake.  I have now reviewed
>>> draft-ietf-lisp-impact-04 and the same comments about security apply.
>>> 
>>> Hilarie
>>> 
>>>> From: Damien Saucez <[e-mail]>
>>>> Date: Tue, 13 Oct 2015 08:13:08 +0200
>>> 
>>> 
>>>> Thank you for the review. I would have a question regarding the document you reviewed. Did you review th
>>> 
>>>> draft-sauces-lisp-impact-04
>>> 
>>>> or 
>>> 
>>>> draft-ietf-lisp-impact-04
>>> 
>>>> Thank you,
>>> 
>>>> Damien Saucez 
>>> 
>>>> On 13 Oct 2015, at 05:01, Hilarie Orman <[e-mail]> wrote:
>>> 
>>>>> Secdir review of LISP Impact
>>>>> draft-saucez-lisp-impact-04.txt
>>>>> 
>>>>> Do not be alarmed.  I have reviewed this document as part of the
>>>>> security directorate's ongoing effort to review all IETF documents
>>>>> being processed by the IESG.  These comments were written primarily
>>>>> for the benefit of the security area directors.  Document editors and
>>>>> WG chairs should treat these comments just like any other last call
>>>>> comments.
>>>>> 
>>>>> A new way of handling routing information has been defined in IETF
>>>>> documents about the Locator/Identifier Separation Protocol (LISP).
>>>>> The draft under discussion here elaborates on the possible
>>>>> consequences of widespread use of LISP.
>>>>> 
>>>>> The draft punts on security considerations and refers to previous
>>>>> documents describing threats to LISP and how LISP uses cryptography
>>>>> for protecting the integrity of its messages.
>>>>> 
>>>>> It seems to me that if the purported impact of LISP is to "scale the
>>>>> Internet", then its impact on security should be a major part of the
>>>>> equation.  Will it make routing information more or less vulnerable
>>>>> malicious manipulation?  How will it affect the stability of a network
>>>>> that is under constant threat of attack?
>>>>> 
>>>>> I don't feel that the draft can achieve its purpose without addressing
>>>>> security.
>>>>> 
>>>>> Hilarie
>>>>> 
>>>>> PS. I was very disappointed to realize that this was not a draft
>>>>> about my favorite programming language.