Re: WGLC for draft-ietf-mboned-driad-amt-discovery

Hi Mikael,

Thanks for the reference, that was interesting reading.

I think part of the problem here is that RFC 7450 already seemed to make
a well-known anyone-can-run-relay discovery address (unless maybe you read
section 4.1.4.2 the right way), and there's a concern that someone might
do that.

I don't think anything in here is meant as advice that anyone should do it.
The goal with this is to make it so the anycast address can be used for
discovery that stays local to an end receiver, and the risk is that someone
might be running one globally even though it's probably a black hole.

I don't think anyone's doing it yet, and if they do maybe we can convince
them to stop.  As long as they don't, I think this usage should be ok, and
would mean there's a check for whether there's a (hopefully local) listener
at that anycast, followed in parallel by finding the sender's relay by
reverse IP lookup, and still failing over if there is a listener but it's
a black hole (as long as the gateway follows the advice from 2.5.4 in DRIAD).

So I don't think this is meant to be asking for an implementation of a
well-known anyone-can-run-relay address.  In this draft, it's only meant for
ISPs trying to offer multicast to their nearby downstream receivers, and it's
not supposed to get propagated elsewhere.

I'm not sure what I can do about possible misinterpretations of RFC 7450, and
people who think they should run it globally.  Do you think DRIAD needs a
RECOMMEND against that?  That might not be a bad idea. 

Thanks and regards,
Jake

On 2019-04-26, 02:17, "Mikael Abrahamsson" <[email protected]> wrote:

    On Tue, 23 Apr 2019, Holland, Jake wrote:
    
    > Suppose we put global anycast after DRIAD, and made the "use a local 
    > relay" advice vague enough that the gateways don't have to use the 
    > anycast address.
    
    I don't know if I misunderstood, but as soon as someone says "anycast 
    address" I get PTSD flashbacks from 6to4:
    
    https://tools.ietf.org/html/rfc7526
    
    Whatever you do, don't implement a well-known anyone-can-run-relay address 
    because it simply doesn't work in real life.
    
    -- 
    Mikael Abrahamsson    email: [email protected]
    

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