How is this related to my point (assuming this was a reply to my message in the first place)? | Artyom Gavrichenkov | gpg: 2deb 97b1 0a3c 151d b67f 1ee5 00e7 94bc 4d08 9191 | mailto: ximaera@gmail.com | fb: ximaera | telegram: xima_era | skype: xima_era | tel. no: +7 916 515 49 58 On Fri, Sep 22, 2017 at 10:18 AM, Randy Bush <randy@psg.com> wrote:
a bit of history for those with short term vision
1995, and large providers were running out of ram to hold the table. sprint was the closest to the edge and falling over; but others were not far behind and could smell the coffee. these were the days where we all intimately knew each others' networks.
nobody's management was gonna pay to upgrade hundreds of routers. sean had to filter to keep from crashing. others, such as asp and i, were also filtering, as much to keep the table down as to protect from idiots such as vinnie from killing us (7007 incident).
so the providers who were concerned and the rirs met at the danvers ietf and agreed to only let /19s and shorter, and swamp space /24s, through if the rirs would please not allocate longer prefixes for a couple of years until routers could be upgraded. rfc 2050 was the result.
eventually, like six yesrs later, customers complained enough that the filters had to be removed. when a big customer or two wanted to get to someone with a /24 in old B space, the filters fell. business wins.
when v4 runout forces folk to put /28s in frnt of nats, the folk with shiny shoes will have a little chat with senior leadership, and they'll cough up the bucks to hold the routes. history repeats.
like the ethernet mfrs tell us that we need to use 4g, 40g, ... instead of 10g, 100g, 1tb, ... life adds zeros.
randy