Re: Classic DHCP over bridge problem

From: Horst Knobloch (horschti2_at_gmx.de)
Date: 04/16/05 

Date: Sat, 16 Apr 2005 02:26:03 +0200

Todd Knarr <tknarr@silverglass.org> wrote:

> In comp.os.linux.networking <d3pb0j$qck$1@ulysses.news.tiscali.de> Horst
> Knobloch <horschti2@gmx.de> wrote:
>> Why should a bridge modify DHCP requests from its bridged
>> devices? Even if it got an IP address (for management
>> purposes) assigned, it's still a bridge which should not
>> modify bridged traffic. I'm sure I've missed something
>> here.
>
> It isn't modifying the DHCP request. It's modifying the Ethernet
> frame that carries the DHCP request.

It shouldn't do this either. The bridge must be
transparently forward the ethernet frames. (Except if
the destination MAC of the frame is on the same ethernet
port the frame came in, then the frame is dropped).

> Specifically, when the bridge
> sends a frame on behalf of a device on the far side, the MAC
> address the DHCP server will see is the MAC address of the bridge,
> not of the device that really sent the request.

Even if a bridge would do this (which it doesn't), it
shouldn't confuse the DHCP server. The DHCP server
looks into the DHCP request for processing and doesn't
bother with the ethernet frame overhead.

This should be the normal behavior but may be the
XP bridging software behaves otherwise and processes
the DHCP traffic in a more "creative" way.

Ciao, Horst 

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

Relevant Pages

  • Re: Classic DHCP over bridge problem
    ... > Why should a bridge modify DHCP requests from its bridged ... > modify bridged traffic. ... It isn't modifying the DHCP request. ... sends a frame on behalf of a device on the far side, the MAC ...
    (comp.os.linux.networking)
  • Re: Classic DHCP over bridge problem
    ... > frame that carries the DHCP request. ... Specifically, when the bridge ... > sends a frame on behalf of a device on the far side, the MAC ...
    (comp.os.linux.networking)