[K12OSN] e1000 client nic problem

dale dalen at czexan.net
Sat Oct 16 02:46:26 UTC 2004


Timothy Legge wrote:
> On Fri, 2004-10-15 at 19:17, dale wrote:
> 
>>Hello,
>>	I am setting up a k12ltsp 4.1.1 server and a few clients at home with 
>>an 8 port Gb switch.  I have an nforce based client with builtin 100Mb 
>>nic.  I added an intel 82546 1Gb nic to net boot the client.  When it 
> 
> 
> Why not just use the nvidia nic?  Unless you just want to play with
> gigabit there is no real need (especially in a small setup).
On this system, I just want to play with gigabit:)  I would like to see 
if it makes a difference with tuxracer.
> 
> 
>>tries to boot, it errors out using forcedeth module (says something 
>>about changing option 129 in /etc/dhcpd.conf).  I suspect the boot 
>>process sees the nforce nic and assumes I want to boot from it rather 
> 
> 
> Yes, on board nics are typically seen first.
> 
> 
>>than the intel nic.  I modified /etc/dhcpd.conf on the server, changing 
>>the mac address for ws001 and option 129 to e1000.  Now when it tries to 
>>boot, it says something about too big (It flashes by fast) and then goes 
>>to a cannot boot screen.  tail -f /var/log/messages only shows tftp 
>>client does not accept options
> 
> 
> In the dhcpd.conf file you probably need to specify pxelinux.0 as the
> filename because it sounds like you are using the e1000's pxe
> implementation.
> 
> Tim
Thanks Tim!  changing the filename to pxelinux.0 for ws001 did help, but 
now I am back to the forcedeth module problem.  I am going to read the 
docs on k12ltsp.org and look at the wiki.

Thanks again,

Dale




More information about the K12OSN mailing list