> On 18 Sep 2016, at 18:20, Massimiliano Stucchi <m...@stucchi.ch> wrote:
> I would suggest moving to head. I spent an inordinate amount of time
> trying to debug issues I was having with 15.05.1, and then found out
> that upgrading to head made the network more robust.
Thanks for that assurance. Is HEAD always stable enough to at least avoid
bricking the router? I note that there’s also some enthusiasm for LEDE, but I
don’t think that all installs of homenet work.
> I have a similar situation, although on a fiber connection that requires
> PPPoE for IPv4. I solved it like this:
> config interface 'homenet4'
> option ifname 'eth0.6'
> option proto 'pppoe'
> option username 'username'
> option password 'password'
> config interface 'homenet4ext'
> option ifname 'pppoe-homenet4'
> option proto 'hnet'
> option mode 'external'
> option _orig_ifname 'pppoe-homenet4'
> Basically, you "piggyback" on the interface created by PPPoE to use it
> for hnet.
That’s pretty much where I got to for pppoe, although I wasn’t (still amn’t?)
confident that I understand what ‘interface’ does within OpenWRT, especially
if/how it stacks up use of different ifname’s within /etc/config/network and
/etc/config/wireless. I have noticed that WAN/LAN and WAN6 have special
meanings: WAN6 is not aware of all protocols, for instance. There are also
oddities such as WAN_6 which seem to appear. I believe that the _orig_ifname
is an artifact of Luci. But I could be wrong.
> Again, I suggest moving to head. You'll find more mature code.
Ok. And clearly hnetd in particular will have moved on.
Any insight into how I can get the hnetd use of dhcp to send the clientid and
vendor class identifiers (options 61 and 60) as these are used by my isp to
establish the connection? I think that they also have some timing constraints
relative to dhcpv6 exchange.
> Massimiliano Stucchi
homenet mailing list