Re: [v6ops] reclassify 464XLAT as standard instead of info

[Alexandre Petrescu <[e-mail]>]

Le 20/09/2017 à 02:51, JORDI PALET MARTINEZ a écrit :
> I’ve deployed 464XLAT in cellular networks with SLAAC but also with 
> DHCPv6-PD using LTE “routers”.

I would like to ask you which LTE modem was inside the LTE router?
Because the Balong modem in some Huawei smartphones, and the some
Qualcomm MDM8xxxx series in some IoT devices block DHCPv6 messages.

Was the LTE network providing a DHCPv6-PD server?

> Same for residential networks.

Residential network is a different matter.

The residential network I live on is not 464XLAT.  And that has some 
users too.

> I don’t think one or the other is relevant to the suggestion I’m 
> doing about reclassifying it as standard.

Why?

> If you do it with SLAAC, as described in the RFC,

Jordi - one cant get a prefix with SLAAC.  If there's some
RFC saying the contrary then that's INFORMAITONAL and should stay that way.

> the NAT46 will be stateful (by means of a previous NAT44)

I guess we dont want either kind of NATx to be any form of StdsTrack.

> because you
> don’t have a specific /64 for the translation. Otherwise will be
> stateless.
> 
> The CLAT, which is the “client” side (run in a cellular phone or a 
> CPE or OS), is available in open source. Just look at the OpenWRT 
> implementation, but I’m sure there are others.

OpenWRT is made for residential networks, not for cellular.

For example, it has an 'odhcp' client very thin which is very good.  But 
it takes time and effort to bring it to cellular modem.  The modem is 
different in cellular vs residential gw.  The size of devices is maybe 
an order of magnitutde different (think set-top-box vs matchbox).

I am not sure which part of OpenWRT (a whole OS distribution) you mean 
for 'CLAT'.  Is there some CLAT command line.

Alex