Simple dhcrelay question

Hi folks,

How does a dhcp server figure out what subnet clause to use for
assigning ip addresses when the originating NIC knows nothing about itself?

Does it work it out from the netmask of the very first router interface
that the dhcp discover hit?

thanks

--
Andy Richardson

We need Linux, like dolphins need a hole in the head.

Andy Richardson <(E-Mail Removed)> wrote:

> How does a dhcp server figure out what subnet clause to use for
> assigning ip addresses when the originating NIC knows nothing about
> itself?
>
> Does it work it out from the netmask of the very first router interface
> that the dhcp discover hit?

No. The dhcp server knows this because this information
was provided by configuration. Google for ISC dhcpd and
have a look to the documentation of the config files.

Ciao, Horst
--
»When pings go wrong (It hurts me too)« E.Clapton/E.James/P.Tscharn

Andy Richardson wrote:
> Hi folks,
>
> How does a dhcp server figure out what subnet clause to use for
> assigning ip addresses when the originating NIC knows nothing about
itself?
>
> Does it work it out from the netmask of the very first router
interface
> that the dhcp discover hit?

Assuming the server is not _on_ the subnet, but using a relay ...

IIRC, it's this (from RFC 951):
http://www.faqs.org/rfcs/rfc951.html
[q]
If a gateway [ie., relay agent] does decide to forward the request, it
should look at the 'giaddr' (gateway IP address) field. If zero, it
should plug its
own IP address (on the receiving cable) into this field.
[eq]

Thus, the server knows on what subnet the request originated.

hth,
prg
email above disabled

Thanks guys
(I didn't mean netmask originally anyway. oops)


Thanks especially for the rfc 951. I've read down to about section 8
and my head was starting to hurt. It all makes good sense though. I'll
try again in the morning.
I like the writing style. I must start visiting RFC's. I knew they
existed but assumed that they would be very dry and stiff.


Thanks again.
--
Andy Richardson

We need Linux, like dolphins need a hole in the head.

dry and stiff ? the rfc's ? you obviously haven't read rfc 1149,
"Standard for the transmission of IP datagrams on avian carriers"

http://www.faqs.org/rfcs/rfc1149.html

Andy Richardson <(E-Mail Removed)> wrote:


> Thanks especially for the rfc 951. I've read down to about section 8
> and my head was starting to hurt. It all makes good sense though. I'll
> try again in the morning.
> I like the writing style. I must start visiting RFC's. I knew they
> existed but assumed that they would be very dry and stiff.

Ok. That's good because then you aren't that frustrated
when I say that you need to read for DHCP the RFC 2131. ;-)

However the good thing is that DHCP is the bigger brother
of BOOTP and they are similar since DHCP is based on
BOOTP. DHCP was designed in a way that it can also serve
BOOTP clients.


Ciao, Horst
--
»When pings go wrong (It hurts me too)« E.Clapton/E.James/P.Tscharn

if dhcp messages were relayed, there will be an option gateway, server
will do longest match. gateway address will be the interface address
where the packet come from.

>dry and stiff ? the rfc's ? you obviously haven't read rfc 1149,
>"Standard for the transmission of IP datagrams on avian carriers"
>
>http://www.faqs.org/rfcs/rfc1149.html

I don't want one of them avian carrier masts near my home. Hell, just
think of all those critters trying to er... "rapidly gain height by
losing ballast" ... over my car windscreen.


> Ok. That's good because then you aren't that frustrated
> when I say that you need to read for DHCP the RFC 2131. ;-)

Bye 'eck , forget reading it fresh in the morning, I'll need a long
weekend to get through that lot. Still, the UK has a bank holiday coming
up - quiz me after that. ;-)


--
Andy Richardson

We need Linux, like dolphins need a hole in the head.