Re: [RRG] Consensus check: mapping granularity

[Brian E Carpenter <brian.e.carpenter@gmail.com>]

On 2008-03-26 06:16, William Herrin wrote:
> On Sun, Mar 23, 2008 at 12:55 AM, Stephen Sprunk <stephen@sprunk.org> wrote:
>> Thus spake "Olivier Bonaventure" <Olivier.Bonaventure@uclouvain.be>
>>  > The identifier to locator mapping function MUST support mapping entries
>>  > for aggregates of identifiers. It MAY also support mapping entries for
>>  > host identifiers.
>>
>>  I cannot imagine how a mapping system could be capable of mapping aggregates
>>  of identifiers without also being capable of mapping individual ones.
> 
> Stephen,
> 
> In a monolithic push system you'll want to aggregate maps as they
> become distant from their origin. When asking whether the mapping
> function must support individual host identifiers, we're really asking
> whether that aggregation is permitted to be mandatory or whether it's
> only allowed to be opportunistic.
> 
> I say the latter: the mapping function MUST support mapping entries
> for individual host identifiers anywhere such is necessary to achieve
> the origin's desired functionality. Otherwise we're back to wasting an
> entire /24 (or whatever the distance aggregate is) in order to support
> a single anycasted server.

Well, that's a side-effect of BGP filtering practice, isn't it?
As far as I can see, filtering simply won't be possible or
meaningful in map distribution, so an anycast address will
normally be mapped 1:n to a set of RLOCs (i.e. will have n
1:1 mappings). It's highly unlikely that any of those
mappings could aggregate with any other prefix mapping to
the same RLOC.

    Brian

--
to unsubscribe send a message to rrg-request@psg.com with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg