ipv6 accepted & announcement size upto /48 or longer than /48 ?


Am i right ?

Policy for ipv4 accept and send upto /24
Policy for ipv6 accept and send upto /48

Internally, sure.


Sending up to a /24 is an oversimplification in today's post-IPv4
exhaustion world. ARIN now has a "Dedicated IPv4 block to facilitate IPv6
Deployment" ( from which they are may allocate in /24 thru
/28 sized blocks.

RIPE has tested visibility of longer than /24's and reported on that in
2014 and 2015, both with and without IRR entries.

Theodore Baschak - AS395089 - Hextet Systems
https://bgp.guru/ - https://hextet.net/
http://mbix.ca/ - http://mbnog.ca/

NTT accepts "longer-than-/48" (and similarly "longer-than-/24") prefixes
only from customers, if an exact matching route6: or route: object
exists in the IRR. NTT will announce such more-specifics to customers,
but not to peers.

This is useful for customers who want to use NTT's backbone to connect
their satellite sites to each other, as an alternative to building out
their own backbone.

Personally I wouldn't expect global reachability for prefixes longer
than /48 or /24, so you'll see that most people make sure a covering
announcement also exists. In a controlled environments these
more-specifics can be very useful.

Kind regards,


Yes, but that's not a policy, that's a BCP.

As an example, the RIPE community has
   documented that, at the time of writing of this document, IPv4
   prefixes longer than /24 and IPv6 prefixes longer than /48 are
   generally neither announced nor accepted in the Internet [20] [21].