Re: [Int-area] RAO for IPv4
Roland Bless <[email protected]>
| Newsgroups | gmane.ietf.nsis,gmane.ietf.int |
|---|---|
| Organization | Institute of Telematics, University of Karlsruhe |
| Message-ID | <[email protected]> |
Hi Thomas, Thomas Narten wrote: > Why exactly is it necessary to have the IPv4 and IPv6 number space be > consistent? What is the technical requirement/benefit here? > > My own sense is that this would be "nice to have", but it is certainly > not required, and I don't see that having the numbering space be strictly > synchronized as providing compelling technical benefit. > > What am I missing? I would say: If the numbers are aligned you don't need separate mapping tables for IPv4 and IPv6. The idea is that you map an NSLP ID, i.e. a an identifier for a specific signaling application like QoS NSLP or NAT/FW NSLP to an RAO value as demultiplexer. In this case you would only need one mapping table and not two different tables for v4/v6. Technically you can have separate v4/v6 RAO tables, but using only one would be easier to implement etc. Roland