[c-nsp] Conditional advertise-map

[Heath Jones]

I thought you'd say that...

There is absolutely *NO* reason why an additional entry needs to occupy
space in TCAM in this case.
If you have 2 contiguous prefixes that can be directly aggregated to a
single prefix on the next subnet boundary, and that share the same
next-stage-treatment (egress, queueing etc), they can take up only 1 spot.
If a router isn't preprocessing routing information before it goes to the
forwarding plane, then its not efficiently using the expensive resources..

I get your point about the bgp table size, but i'm not suggesting he breaks
a /16 into /24's or something... Please keep that in mind! I am well aware
of the point you are making, but still believe that my suggestion is a
viable solution to the problem.

Oh, and for the record, it would change from 1 prefix into 3, not into 2
like i suggested before (my fault).


On 15 September 2010 18:28, Gert Doering <gert at greenie.muc.de> wrote:

> Hi,
>
> On Wed, Sep 15, 2010 at 05:55:24PM +0100, Heath Jones wrote:
> > You will probably find that the as path prepending will chew more memory
> > than 1 more prefix matching an existing as path so the subnetting option
> > works out better.
>
> To the contrary.  Different AS paths are "just DRAM".  Every additional
> prefix is "forwarding memory", read: limited and very expensive TCAM
> (or SRAM or whatever).
>
> [Lower-end routers have BGP tables and forwarding tables in the same
> DRAM memory, so there it's "just DRAM", but for higher-end, the forwarding
> tables go to hardware memory]
>
> > just not cricket. I'm just trying to help him out with some suggestions.
> It
> > would turn 1 prefix into 2 - thats hardly unreasonable..
>
> Right now, about half(!) of the global routing table is deaggregates.
>
> If those people would all aggregate properly, the routing table would
> still fit into non-XL TCAMs - for us alone (and we're a small ISP) that
> would have saved about 150.000 US$ in not-yet-needed TCAM upgrades.
>
> It *does* matter.
>
> gert
> --
> USENET is *not* the non-clickable part of WWW!
>                                                           //
> www.muc.de/~gert/
> Gert Doering - Munich, Germany
> gert at greenie.muc.de
> fax: +49-89-35655025
> gert at net.informatik.tu-muenchen.de
>