Hi Joel,

On 3/5/15 5:37 AM, Joel M. Halpern wrote:
> I am not sure what would go in the introduction document that would
> address Radia's concern.
> 
> Each mapping system document includes security considerations that are
> specific to that mapping system.  The Original BGP based system has one
> set of properties, the DDT one (which we will be advancing after the
> blocking documents are cleared) as a different set.  The introduction
> document is not a requirements document driving that work, but rather an
> explanation.

Agreed.

> I suppose we could add a short paragraph that notes that individual
> mapping systems need to address traffic misdirection threats, and that
> the mapping system interfaces (Map Servers and Map Resolvers, and the
> protocol between ITRs/ETRs and those components) need to address a
> multitude of security issues.  But would that kind of text really help
> the introduction?
> 

I think it would help by letting readers know that there are different
security considerations for the different types of mapping systems.

> I very much don't want to see this document getting into the question of
> why not to use DNS as the mapping system, or a DNS clone, or ...  One of
> the keys to the LISP design is that different mapping system designs can
> be and are being tried.  One of the challenges in writing this
> introduction is to reflect the existing RFCs while allowing for the new
> work such as DDT that we know the working group plans so as not to
> obsolete this introduction.

I agree that this is not the document to get into the "use vs. not use"
the DNS for mapping.

Regards,
Brian

Attachment: signature.asc
Description: OpenPGP digital signature

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

Reply via email to