[RAM] Re: Ramblings about "locator"

Earlier, Brian Carpenter wrote:
> I was thinking about why we're having trouble being crisp
> about "locator".
>
> Maybe part of it is that in some contexts, as Noel hinted,
> a locator is loaded with topological significance. If the
> network topology in a given region is a strict binary
> hierarchy, then the locator may be very tightly mapped
> to the topology (and is in effect a route). But in other
> regions of the network where the topology is an arbitrary
> graph, the same locator has no mapping to topology and needs
> to be bound to a route by a routing protocol.

A locator always has topological significance, IMHO.
However, a locator's topological significance is not
necessarily bound to a hierarchical bit-pattern
(though with IP routing the address does have that
sort of binding).

> Another thought is that on a classical broadcast Ethernet,
> an Ethernet address is a locator, but a rather strange one
> in that nobody except the receiver knows where the located
> interface is. But the same Ethernet address on a switched
> Ethernet or in a spanning tree becomes much more
> like a network-level locator; it's mapped to topology by
> the switching or bridging mechanism.

	As someone who builds Ethernet switches these days,
my take is that the IEEE MAC address (either 802 or 1394)
in an Ethernet frame is an address (i.e. neither identifier,
nor locator).

A typical Ethernet switch simply has a table inside.  One feeds
in a destination MAC address and gets back which (bridge/switch)
interface(s) to send the frame out.  If the target MAC is not
in the table, one sends the frame out all interfaces.

{The claim that "an IEEE MAC is an Identifier" would be more
true if one considered Ethernet-II rather than IEEE 802.3.
Ethernet-II said there was one MAC address per node, while
802.3 has one MAC per interface.  Both definitions can work
fine with off-the-shelf bridging algorithms.}

> Maybe the essential point is that a locator can at least in
> principle be mapped to topology and an identifier can't.

I would phrase it differently.  A locator has location semantics,
but not identity semantics.  An identifier has identity semantics,
but not location semantics.  An address has both kinds of semantics.
So an MAC address in an Ethernet frame is an address.

And by the way, I view an EUI-64 (e.g. as used in O'Dell's
proposals) to be an Identifier, neither an address nor a locator.

Ran

_______________________________________________
RAM mailing list
RAM@iab.org
https://www1.ietf.org/mailman/listinfo/ram