Re: [Model-t] [e-mail] list description

[Michael Richardson <[e-mail]>]

Stephen Farrell <[e-mail]> wrote:
    >> There are often different situations for LLNs, Data-Centers, and the
    >> like.

    > Sure. Those are not IMO out of scope at all. But neither is any one of
    > those the full scope - I think the scope we ought consider here is
    > basically any network technology that the IETF does/controls/specifies
    > and even a bit more than that, as IETF-defined things can of course be
    > attacked from lower layers or physically etc.

What I'm trying to say is that there are some threats that we deal with
on the Capital-Internet that are far more manageable in the small.

We then get into trouble when someone takes a DC technology and uses it
across the Internet, or when the DC becomes multi-tenant, etc.

    >> Is the point of this list to talk about threats on the Internet, or
    >> about better explaining threat models assumed in documents, such that
    >> we have better Security Considerations?  The tasks are not mutually
    >> exclusive.  Being better able to say, "mechanism X deals with the
    >> Internet Threat model situation Y" is a great win for Security
    >> Considerations.
    >>
    >> At first reading, I thought it was more about articulating threat
    >> models better, but upon further reading, I think this list is more
    >> about specific threats.

    > I'm not sure I'm quite getting what you mean exactly. FWIW, I do think
    > we'll be discussing threats on this list, but a useful output will be a
    > distillation of that discussion that gives us the kind of win for
    > security considerations you have in quotes above.

For instance, https://datatracker.ietf.org/doc/rfc7416/ provides a list of
threats to LLNs.   Having named them, I don't have to repeat them in other
documents.  That's really useful.

AFAIK, we don't (yet?) have an IETF document that details the problem with
predictable/guesstable IDs, such as DNS query IDs or TCP SYN numbers.
{yes, Fernando has a document that goes that way}.  Does that threat have a name?
Does the name change based upon whether the attacker is on-path (able to
observe traffic), or off-path?  Naming such things would be useful.

--
]               Never tell me the odds!                 | ipv6 mesh networks [
]   Michael Richardson, Sandelman Software Works        |    IoT architect   [
]     [e-mail]  http://www.sandelman.ca/        |   ruby on rails    [





--
Michael Richardson <[e-mail]>, Sandelman Software Works
 -= IPv6 IoT consulting =-