Re: [sidr] WGLC: draft-ietf-sidr-bgpsec-reqs

[Randy Bush <[e-mail]>]

>> hence the "per se," meaining in and of itself.  some cases of pouring
>> cement into a router (see london tube) are security issues, some are
>> not.
>>
>>how would you make that more clear?
>
> I think Warren’s suggestion of simply eliminating the assertion about
> whether it’s a security issue, per se or otherwise, and just saying
> that it’s out of scope is enough for the intro.

i disagree.  would be interested in hearing other opinions.

>> they are announcements of P by A to B which are not agreed by all
>> parties concerned (including A, B, neighbors of A and B, the
>> originator of P, ...).  the problem lies in detecting them,
>> especially from a distance.
> So I think that goes back to my suggestion that since you already
> discuss intent in 3.22, that might be a place to add something about
> leaks, either as a part of that req or a follow-on, because that’s
> really what you’re saying here - we understand theoretically what they
> are, but not how to detect them such that we could do anything to
> prevent the undesired ones.

leave that to grow/idr.  this is routing security.

randy