[gPXE] arp proglem

[何闯]

Dear all,


By debug with parameters DEBUG=ipv4,netdevice, I finally found that the delay is because gPXE is busy polling packets from other clients, which is not its own business(ipv4 say: rejected packets, not connected). That's why gPXE can not hand ARP request immediately. So
my question is that, how can I avoid this problem and let the gPXE handle ARP request ASAP?




yours,
soforth

At 2011-04-12，"何闯" <justhechuang at 163.com> wrote:

Dear all,
 
This problem is as bellow:
192.168.1.3 and 192.168.1.2 are two bootloader client, form some reason, I need them to communicate with each other via UDP protocol.
However, 1.3 to 1.2 UDP packet invokes ARP transsmision,FIVE seconds have past before 1.2 receive 1.3's ARP request, resulting in a UDP sending timeout.

192.168.1.5 is dhcp & tftp server, seems that 1.5 communicates with both 1.3 and 1.2 are OK. What's wrong here?
 
 
yours,
soforth
 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://etherboot.org/pipermail/gpxe/attachments/20110412/1e68b3ce/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 111.jpg
Type: image/jpeg
Size: 134081 bytes
Desc: not available
URL: <http://etherboot.org/pipermail/gpxe/attachments/20110412/1e68b3ce/attachment-0002.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 1.jpg
Type: image/jpeg
Size: 100134 bytes
Desc: not available
URL: <http://etherboot.org/pipermail/gpxe/attachments/20110412/1e68b3ce/attachment-0003.jpg>