Re: [v6ops] draft-ietf-v6ops-ula-usage-recommendations - work or abandon?

[Owen DeLong <[e-mail]>]

> On Nov 12, 2015, at 17:04 , Fernando Gont <[e-mail]> wrote:
> 
> On 11/13/2015 03:05 AM, Owen DeLong wrote:
>>>> They still work fine on networks that
>>>> operate the way the Internet was originally designed with end-to-end
>>>> connectivity.
>>> 
>>> not sure what you mean by "end-to-end" connectivity.
>> 
>> I won’t speak for Lorenzo, but what I mean when I talk about “end-to-end”
>> is the situation where the L3 packet that leaves the origin host arrives
>> at the destination host without being modified except for certain fields
>> intended to be modified in flight. (Namely TTL/Hop Count, L2 and lower
>> layer fields, etc.). Certainly the L3 and higher layer identifiers are not
>> to be tampered with in flight.
> 
> FTP PORT doesn't work fine in firewalled environments. It requires a
> middlebox to tap into the FTP control protocol so that the fw can open
> the necessary holes. -- So the only networks where FTP PORT works fine
> is un-NATed, unfiltered networks -- which is kind of Wonderland.

FTP is not the only protocol being discussed and there are actually a variety
of alternative ways to solve the FTP PORT problem that do not require
removing the firewall or having an unfiltered network.

Owen