Re: [RAM] Ramblings about "locator"

[Stig Venaas <stig.venaas at uninett.no>]

Brian E Carpenter wrote:
> On 2007-06-15 05:03, Ved Kafle wrote:
>>> Marshall,
>>>
>>> ...
>>>>> Maybe the essential point is that a locator can at least in
>>>>> principle be mapped to topology and an identifier can't.
>>>>>
>>>> If in Ethernet what is essentially a random number (the MAC 
>>> address)
>>>> becomes
>>>> a locator then maybe the distinction will never be crisp.
>>> Well, that horrible thought crossed my mind. I think I disagree
>>> with Jarno's comment, because as soon as an Ethernet address
>>> gets into a spanning tree it has become a locator (at level 2).
>>
>> So, the distinction between an identifier and a locator is dependent on
>> the avalaibility of routing/switching infrastructure. If networks
>> possess routing/switching insfrastructure based on a
>> identifier namespace, the identifier simultaneously functions as a
>> locator. 

Yes, I would say that it functions as a locator, but I'm not sure I
would say that it is a locator. Does the fact that you know where all
the identifiers are located make them locators?

> My original RAMble (sorry) ended thus:
> 
>  Maybe the essential point is that a locator can at least in
>  principle be mapped to topology and an identifier can't.

So I'm not sure exactly what "mapped to topology" should mean, but I
think I do. I would not say that MAC addresses can be mapped, their
values are not tied to the topology in any way.

Stig

> I think that's correct and actually consistent with much that's been
> said. A slightly different way to say it is that a locator is a
> handle for a route.
> 
>     Brian
> 
> _______________________________________________
> RAM mailing list
> RAM at iab.org
> https://www1.ietf.org/mailman/listinfo/ram


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