Re: [Pidloc] Encoding Routing in FAST

[Tom Herbert <[e-mail]>]

On Fri, Oct 12, 2018, 7:56 AM Behcet Sarikaya <[e-mail]>
wrote:

> Hi all,
>
> Tom Herbert published a draft
>
> https://tools.ietf.org/html/draft-herbert-route-fast-00
>
> which describes a solution by which the client (mobile node) can send its
> locator to the sender.
>
> Tom defined the locator encoding for ILA and the network seems to be
> specific to ILA.
> However I believe this design can be generalized to other IdLoc protocols
> or at least the locator encoding could be reused..
>

Hi Behcet,

Thanks for posting and making connection.

Yes, ILA is an example. Tickets can contain other types of locators or
information.

>
> Basically this draft represents a solution (ILA based) for the locator
> sharing cases mentioned in Erik's draft.
>

It might be a little different in that regard. FAST could encode locator
for return path in a ticket. Peer just reflects ticket and doesn't know
about locator, it's used to route packets in destination provider network.
So locators aren't actually being shared, they're transparently carried by
third parties.

Actual sharing could be done, for instance a locator as the ingress point
in a network might ambiguous enough to avoid privacy concerns. A ticket
might just contain locator (and network overlay method). In this case peers
would need to parse tickets, so the might go down path of a standardized
ticket format.

Tom




> Any comments?
>
> Behcet
> --
> Pidloc mailing list
> [e-mail]
> https://www.ietf.org/mailman/listinfo/pidloc
>