Re: route policy (Re: Public shaming list for ISPs announcing other ISPs IP space by mistake)
Yes, RIPE rock. Please make it all not suck.
Unfortunately, RIPE DB will allow anyone to add any route objects for prefixes that are not under the RIPE management :-(. For example, anyone could add route objects for most of DNS root server prefixes.
Little details get used to avoid fixing bigger problems, see the please stop sucking bit. If there was somewhere else the aliens had to be registered then they could be dropped from RIPE brandon
On Fri, 15 Aug 2008, Brandon Butterworth wrote:
Yes, RIPE rock. Please make it all not suck. Unfortunately, RIPE DB will allow anyone to add any route objects for prefixes that are not under the RIPE management :-(. For example, anyone could add route objects for most of DNS root server prefixes.
Little details get used to avoid fixing bigger problems, see the please stop sucking bit.
If there was somewhere else the aliens had to be registered then they could be dropped from RIPE
I'm not sure I follow. Many of these aliens are in fact registered in RADB, so AFAICS, there that is no reason for them to be registered in RIPE DB. On the other hand, some want to register them in RIPE DB because some operators just want to use RIPE DB e.g. for data consistency etc. reasons. But putting data without practically any authorization in RIPE DB doesn't seem to be a useful model in the long run. -- Pekka Savola "You each name yourselves king, yet the Netcore Oy kingdom bleeds." Systems. Networks. Security. -- George R.R. Martin: A Clash of Kings
I'm not sure I follow. Many of these aliens are in fact registered in RADB, so AFAICS, there that is no reason for them to be registered in RIPE DB.
when ripe will not mirror the irr segment in which they do register. randy
On Fri, 15 Aug 2008 13:56:09 +0300 (EEST), Pekka Savola wrote:
I'm not sure I follow. Many of these aliens are in fact registered in RADB, so AFAICS, there that is no reason for them to be registered in RIPE DB.
On the other hand, some want to register them in RIPE DB because some operators just want to use RIPE DB e.g. for data consistency etc. reasons. But putting data without practically any authorization in RIPE DB doesn't seem to be a useful model in the long run.
As I understand things, the "without practically any authorization" model holds for *everything* registered in the RADB. Right? If that's not a useful model for the RIPE DB, what about the RADB? --Sandy P.S. Not to pick on the RADB. Most IRRs, as I understand it, enforce little in the way of authorization. It's just that the RADB was mentioned.
participants (4)
-
Brandon Butterworth -
Pekka Savola -
Randy Bush -
sandy＠tislabs.com