1. "DHCP Request" is sent out from "eth0.0" of a Linux box. It's a VLAN tagged packet with VID set to 0.
2. "DHCP Offer" is sent back from the switch and received by Linux. It's a regular Ethernet packet without any VLAN tag. Since not tagged, it is sent to interface "eth0" instead of "eth0.0 ". Then the packet gets dropped by "eth0" because the DHCP client is started on "eth0.0".
I don't think it's a configuration problem on the switch or DHCP server. Please correct me if my assumption is wrong.
But if it's a configuration problem on Linux, I'd be happy to know what I should do.
On Dec 13, 2007 12:00 PM, <vlan-request@xxxxxxxxxxxxxxx
> wrote:
Date: Thu, 13 Dec 2007 01:58:45 +0100
From: Peter Stuge < stuge-vlan@xxxxxxx>
Subject: Re: 802.1p only?
To: vlan@xxxxxxxxxxxxxxx
Message-ID: <20071213005845.1715.qmail@xxxxxxx >
Content-Type: text/plain; charset=us-ascii
On Wed, Dec 12, 2007 at 02:37:16PM -0800, Xin Huang wrote:
> Here is the problem: when a DHCP request is sent out of eth0.0, the
> switch sends back a DHCP offer packet without any VLAN header.
So your switch is the DHCP server?
This all seems like a configuration issue.
> But is there a simple way to make DHCP client work on eth0.0?
You wrote above that your DHCP client is already communicating over
eth0.0. If in fact it is not, this would be rather simple, depending
on your client software:
dhcpcd eth0.0
dhclient eth0.0
pump -i eth0.0
etc.
//Peter
------------------------------
_______________________________________________
Vlan mailing list
Vlan@xxxxxxxxxxxxxxx
http://www.candelatech.com/mailman/listinfo/vlan
End of Vlan Digest, Vol 34, Issue 1
***********************************
_______________________________________________ Vlan mailing list Vlan@xxxxxxxxxxxxxxx http://www.candelatech.com/mailman/listinfo/vlan
