Re[2]: whats the deal with 54fx:: ?

[Bob Fink]

Michael,

At 01:01 AM 4/30/2002 +0200, Michael Kjorling wrote:
>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>On Apr 29 2002 20:25 +0200, Sascha Bielski wrote:
>
> > BM>         hijacked address space?  54f8:: is not delegated.
> >
> > iirc that was some "old" testbed. still many 5xxx::/16 records in
> > 6bone whois database...
>
>I seem to recall having read something about that, somewhere - but
>it's got to have been a long time ago, because there is no way you can
>squeeze 5000::/4 (format prefix %010 == 4000::/3) into the IPv6
>globally aggregatable address range (which has a format prefix of
>%001, making for 2000::/3 - anyone recognizes that range?)

I'll cover the history here for the record for 5F00::/8 prefixes, but will 
note that 54F8::/16 isn't covered by this history, so Bill is correct in 
saying it is hijacked (even if by accident :-).

The original and now obsolete IPv6 Provider-Based Unicast address format 
was specified in RFC2073:

<ftp://ftp.isi.edu/in-notes/rfc2073.txt>

It specified a 4000::/3 prefix as follows based on the then IPv6 Address 
Architecture in RFC1884 (which even reserved some space for 
Geographic-Based Unicast Addresses :-):

>       | 3 |  5 bits  |   n bits   |   56-n bits  |        64 bits     |
>       +---+----------+------------+--------------+--------------------+
>       |010|RegistryID| ProviderID | SubscriberID |  Intra-Subscriber  |
>       +---+----------+------------+--------------+--------------------+

Then a Testing Address Allocation was made in RFC1897 for the 6bone:

<ftp://ftp.isi.edu/in-notes/rfc1897.txt>

which specified 5F00::/8 as follows:

>    | 3 |  5 bits  |  16 bits | 8 |   24 bits  | 8 | 16 bits|48 bits|
>    +---+----------+----------+---+------------+---+--------+-------+
>    |   |          |Autonomous|   |    IPv4    |   | Subnet | Intf. |
>    |010|  11111   |  System  |RES|   Network  |RES|        |       |
>    |   |          |  Number  |   |   Address  |   | Address|  ID   |
>    +---+----------+----------+---+------------+---+--------+-------+

The Provider-Based Unicast address format was then replaced by the IPv6 
Aggregatable Global Unicast Address Format which is specified in RFC2374:

<ftp://ftp.isi.edu/in-notes/rfc2374.txt>

It specified a format as follows, which is part of the revised IPv6 Address 
Architecture in RFC2373 (and currently being updated as we speak):

>      | 3|  13 | 8 |   24   |   16   |          64 bits               |
>      +--+-----+---+--------+--------+--------------------------------+
>      |FP| TLA |RES|  NLA   |  SLA   |         Interface ID           |
>      |  | ID  |   |  ID    |  ID    |                                |
>      +--+-----+---+--------+--------+--------------------------------+

Then a new IPv6 Testing Address Allocation was made in RFC2471 for the 
6bone (replacing RFC1897):

<ftp://ftp.isi.edu/in-notes/rfc2471.txt>

which specified the well known 3FFE::/16 format.

There was a small amount of activity under the old 5F00::/8 format through 
1998. After that time it was supposed to go away, and mostly has, although 
a tiny bit of these 5F prefixes pop up from time to time.


Hope this at least recovers some history rapidly being forgotten.

Bob