On Tue, Nov 11, 2008 at 03:03:19PM -0800, Brian Keefer wrote:
> On Nov 11, 2008, at 2:01 PM, Administrator wrote:
>
>> Brian Keefer wrote:
>>> On Nov 11, 2008, at 12:42 PM, Administrator wrote:
>>>> Nope, didn't help. There must be some other mistery. Now it stops  
>>>> at DHCPOFFER part.
>>>>
>>>>
>>>> DHCPDISCOVER from 00:50:18:48:cb:3d via vlan51
>>>> DHCPOFFER on 192.168.51.3 to 00:50:18:48:cb:3d via vlan51
>>>> DHCPDISCOVER from 00:50:18:48:cb:3d via vlan51
>>>> DHCPOFFER on 192.168.51.3 to 00:50:18:48:cb:3d via vlan51
>>>>
>>>> Any ideas?
>>> Do you have the ability to test on -current?  You might try that.   
>>> Also definitely post a follow-up to Misc@ and Cc:  
>>> [EMAIL PROTECTED] to see if he has any ideas.  I'm not a DHCP  
>>> guru, unfortunately.
>>> He's probably going to need some tcpdump samples to see what options 
>>> are getting passed.  This is what was requested last time:
>>> please include "tcpdump -eni<interface> -Xvvs9999 port 67 or port 68"
>>
>> Ok, I will try -current tomorrow. Do I have to recompile world or just 
>> dhcpd? Will this be enough?
>>
>> # cd /usr/src/usr.sbin/dhcpd
>> # make obj && make && make install
>>
>
>
> For -current you should install a snapshot and go from there.  I believe 
> you can't just update dhcpd because there have been library changes.
>
> Hopefully you have a box you can test on.  I tend to use VMs for this  
> kind of thing.
>
> --
> bk

You should be able to just get -current /usr/src/usr.sbin/dhcpd/options.c
and recompile on your system. The library problem was with my
compiling on a -current system and someone trying to that executable
it on a -release system.

.... Ken

Reply via email to