On Apr 3, 2009, at 12:55 AM, Tony Li wrote:
Hi Dow,
locator A locator is a name that has topological sensitivity at a
given layer and changes if the point of attachment at that
layer changes. By default, a locator refers to layer 3.
It is also possible to have locators at other layers.
Locators may have other properties, such as their scope
(local or global (default)) and their lifetime (ephemeral
or permanent (default)).
It is this "point of attachment at that layer" phrase that is still
problematic. If a host disconnects from one place on an ethernet
segment and connects to another, it can still be reachable without
changing it's IP address. Are you defining this as not changing
the point of attachment (in terms of layer 3)?
_Exactly_. If you change ports on your local Ethernet switch, then
you've only changed the L2 topology, and done nothing at all at
layer 3.
I am ok defining things this way. To ensure we're on the same page,
this means that only the network part of an IP address (from the
perspective of the host) can ever be considered a locator, right?
Also, since the mask that is in effect (for forwarding) at various
points in the topology differs, then the part of the IP address that
we can call a locator also differs at various points in the topology.
Also, to say that that a locator changes when the point of
attachment changes depends on the nature of the routing system.
The locator could stay the same, and there could be a change to the
routing state instead. So there are some "relative topological
scope" types of considerations in "this locator-ness" check, but
the definition above does not capture those.
I'd claim that by definition, that token then isn't a locator. It
sounds again like a forwarding system that's doing flat routing on
an identifier.
I think I agree - this is related to the comment above.
I think that this is mostly a matter of semantics. Since our
concern is scalability, I don't think that flat routing is really
all that interesting of a case anyway.
I am trying to get at this property of "relative flatness" of various
names and/or parts of the IP address at different points of the
topology, even if we are only talking about familiar, aggregation
based approaches to scalability.
In other words, if we say that:
(a) on an ethernet segment (that maps to only a single prefix in the
IP routing system), only the network part of the IP address of
connected hosts is considered to be a (L3?) locator, then
(b) a similar property exists closer to the "middle" of the topology/
routing system if aggregation is used.
In (b), more specific prefixes "hidden behind" the aggregate are
topologically flat from the perspective of processes that are only
aware of the less-specific, aggregate mask.
I think maybe you are accounting for all of this relative-ness with
the phrase "attachment point at that layer"?
R,
Dow
_______________________________________________
rrg mailing list
[email protected]
http://www.irtf.org/mailman/listinfo/rrg
