Re: [LONG] ppp-udeb: A succesful installation using it, some ideas and some requests for advices
eddy.petrisor@gmail.com wrote:
>> No. DHCP and PPPoE support are not related.
>Err, this is a problem in the d-i intergration. D-I should prompt as
No, it's not. You use DHCP to configure the ethernet interface and then
use PPPoE for the ppp0 interface.
>few times as possible. The most usual/convenient case is to have a
>dhcp server on the network and to have all configuration information
>from that dhcp server. Currently D-I asks for manual configuration
>information in case DHCP fails. AFAICS, with the introduction of PPPoE
>configuration, manual configuration should be requested in case the
>PPPoE configuration fails, too (after the automatic identification of
>concentrators).
No, it should always be requested and *before* any attempt to configure
PPPoE.
>> > - if concentrator detection fails, drop to manual configuration
>> If no concentrator is detected then the interface does not support
>> PPPoE and there is no need or reason to configure anything.
>That is if you assume that I ment "manual PPPoE configuration", but I
>ment "static configuration" :-) of the Ethernet card.
Then it's something not related in any way to PPPoE support.
When you start probing for PPPoE the ethernet interface should already
have been configured.
>Is it possible to have a PPPoE configuration ran as in the following
>scheme:
> - configure the Ethernet card through DHCP
> - configure a ppp connection (PPPoE) over the Ethernet card
>configured at the previous step; the nameserver and default route are
>set through the PPP connection?
This is what I have been explaining you need to do.
>I was under the impression that on a PPPoE connection one will have
>always an Ethernet card (ethX) which is just up, but has no IP, and a
>pppY virtual interface which has an IP which is given by the PPP
>server. Is this picture not correct?
Yes. I already explained this multiple times. An IP address is not
*needed* on the ethernet interface, but always configuring one is
simpler and may prevent problems later.
>> >is, when should the PPPoE configuration be attempted? It seems logical
>> After DHCP.
>After DHCP fails, right? Or?
After DHCP or manual configuration of the ethernet interface.
>> >In case the PPPoE connection fails, should the user be given the
>> >chance to configure the card as static? Note that if the concentrator
>> These things are not related.
>They are from a D-I usability POV.
As I explained, they are not.
>> >There is small issue, if the concentrator is searched on a card, then
>> >the card in question must be "ifconfig CARD up" with no IP, should
>> >this be restored if the concentrator is not found? Would it affect the
>> Wrong, the ethernet card should get a proper IP address. Network
>> interfaces used for PPPoE do not need any special configuration.
>I'm afraid your statements sound conflictive with each other.
>What I know is what happens here. The configuration supported by my
>provider works if I "ifconfig eth9 up" with the ethernet not having
>any IP and then start pppd so ppp0 is started.
You do not *need* an IP address on the interface but *should* have one.
And as you noticed, always configuring it is simpler.
>> >How can the configuration information be copied to the target system
>> >after the system is installed?
>> The udev script already does this.
>That is not possible because of various reasons;
It is if you do it correctly.
>I tested an installation and:
>1) the ppp related stuff was not installed by default
It is supposed to be.
>2) the configuration of pppoe on the target system was not the same as
>the one in the d-i envronment.
I do not know what this means, but the d-i configuration will work in
the installed system as well.
>3) it is not possible to copy the pppoe configuration on the target at
>the time the ppp-udeb postinst is ran because the system does not have
>yet configured the destination partition for /. (It is necessary to
Obviously you need to do it later.
--
ciao,
Marco
Reply to: