On Mon, Dec 5, 2011 at 9:46 AM, Frank Ellermann
On 5 December 2011 04:27, Cameron Byrne wrote:
[they = the IETF]
they underscored that point by not rejecting various past attempts at
expanding private ipv4 space like 240/4.
Sorry. S/not rejecting/rejecting/
ACK. The last state I'm aware of is that the 240/4 addresses minus one
were and still are (RFC 5735) reserved for IETF experiments, did I miss
some newer IETF consensus about this?
The addresses, AFAIK, are still in a "no mans" land.
I went on a short-lived quest to make these addresses usable in 2008,
because in 2008, they were not usable and i needed addresses.
Meaning, Linux, FreeBSD, Windows would not accept these addresses in
configuration and Juniper and Cisco router would not only not accept
these addresses as part of their configuration, they would not route
the addresses in transit. Some of this may have changed, but not
enough to make this a clear win.
I was told, by a large vendor of network gear, an IETF direction must
be made to define a purpose for these addresses, like:
Both failed to gain support (i assume), and thus nothing happened. My
assumption is these drafts were killed as "IPv4 life support"
RFC 5735 leaves the use of 240/4 undefined ... it could be used for
public, private, multicast, some future use we never thought of,
carrier pigeons ...
Thus, my feeling is that the IETF implicitly said "no ipv4 life
support by expanding private addresses, the cost of ipv4 will go
higher and higher, we can all see it like a slow moving train wreck,
make your strategies wisely". Making this allocation for draft-weil
is changing the rules, slowing the train wreck, going backwards of
previous guidance(IPv6 is the answer to IPv4 exhaust), while at the
same time increasing the amount of of damage.
Ietf mailing list