PXE-E11: ARP Timeout

Question

I am trying to do a PXE boot from a LTSP server connected directly to the client computer.

My problem is, during netboot, it says:

CLIENT MAC ADDR: 8C 89 A5 E0 6C 86 GUID: 000000000-0000-0000-0000-8C89A5E06C86
CLIENT IP: 192.168.1.7 MASK: 255.255.255.0 DHCP IP: 192.168.1.2
GATEWAY IP: 192.168.1.1

PXE-E11: ARP timeout

PXE-E11: ARP timeout 
PXE-E38: TFTP cannot open connection
PXE-M0F: Exiting PXE ROM.

Reboot and select proper Boot device
or Insert Boot Media in selected Boot device and press a key

I don't understand how to solve this issue. I could boot it up properly earlier, and when the server rebooted, I am getting this error.

Please help!

EDIT:

my /etc/ltsp/dhcpd.conf file is as follows:

option domain-name-servers 8.8.8.8 8.8.4.4;
default-lease-time 86500;
max-lease-time 6059934;
authoritative;
option domain-name "BLAX";
allow booting;
allow bootp;
next-server 192.168.1.2;
filename "ltsp/amd64/pxelinux.0;
ddns-update-style none;
subnet 192.168.1.0 netmask 255.255.255.0 {
    range 192.168.1.3 192.168.1.250;
    option domain-name "fdf.com";
    option domain-name-servers 8.8.8.8;
    option broadcast-address 192.168.1.255;
    option routers 192.168.1.1;
    option subnet-mask 255.255.255.0;
    option root-path "/opt/ltsp/amd64";
    if substring( option vendor-class-identifier, 0, 9 ) = "PXEClient" {
        filename "ltsp/amd64/pxelinux.0;
    } else {
        filename "ltsp/amd64/nbi.img";
    }
}

Network configuration:

/------------\          /-------------\
|            |          |             |
|192.168.1.7 |          |  192.168.1.2|      /------------\
| Client     o----------o   Server    o------o  Router    |
| Machine    |          |  Machine    |      |192.168.1.1 |
|            |          |             |      /------------\
/------------\           /------------/

Is your TFTP server up? And is it the same server you specified in your DHCP options? — Michael Hampton, Aug 01 '13 at 12:43
yes the TFTP server is up - tested it with `$tftp localhost` and then a `get filename` which said `filename does not exist`. So seems to be up @MichaelHampton — , Aug 01 '13 at 12:44
"Connected directly to client computer" What does it mean? And also do you see MAC of client in ARP table on the server? — Danila Ladner, Aug 01 '13 at 13:15
@DanilaLadner it means, I have connected the one lan cable from one PC to another directly, without a switch or hub or router — , Aug 01 '13 at 13:16
@DanilaLadner how do I see the MAC of the client in the ARP table on the server? — , Aug 01 '13 at 13:20
how come then if you are interconnecting 2 pcs directly default getaway is 1.1? what is on 1.1? — Danila Ladner, Aug 01 '13 at 13:22
I have two LAN cards on the server machine. One connects to a router, at 192.168.1.1 which is why I cannot connect to the network with 1.1 - but, the other lan connection is connected to the client machine. — , Aug 01 '13 at 13:24
@DanilaLadner I have drawn a crude diagram of the network - updated in the question. — , Aug 01 '13 at 13:29
I could solve it a little, but a laborous way: I swapped the two network connections, and made `dhcpd` listen to only one ethernet.. this way I could remove that error. But this seems to be only a temporary solution - and will have to do it on each server boot — , Aug 01 '13 at 13:36
Right well, in your setup ARP reply most likely gets sent through 1.1 from the server and never gets to your directly connected host, you can try to see it by running tcpdump on the interface connected to router, you could add static route to 1.7 use second nic interface. — Danila Ladner, Aug 01 '13 at 13:41

score 4 · Answer 1 · edited Oct 13 '15 at 20:49

4

In dhcpd.conf, make sure the lines "option domain-name-servers" and "next-server" have the correct IP address.

In the /var/lib/tftpboot/linux-install/pxelinux.cfg files, make sure root=nfs:IP:/directory has the correct IP address.

edited Oct 13 '15 at 20:49

Kara

105
7

answered Jan 09 '15 at 16:40

Chuck

41
2

score 1 · Answer 2 · answered Aug 26 '14 at 05:37

I had such problem when forgot to edit DHCP server configs after changing IP addresses for infrastructure (for the DHCP server too). To fix that:

sed -i -e 's/old_IP/new_IP/g' /etc/dhcp/dhcpd.conf
service dhcpd restart

In addition, the PXE server menus should be updated. In my case:

sed -i -e 's/old_IP/new_IP/g' /*/pxe_server/tftproot/*/pxelinux.cfg/*

score 0 · Answer 3 · answered Aug 24 '16 at 12:55

0

I also had this issue with a PXE-11 ARP timeout on warewulf boot. For anyone else who is dealing with the issue, check your switch if the other posted solutions don't work. My problem was a bad gigabit switch.

answered Aug 24 '16 at 12:55

Patrick

101
1

score 0 · Answer 4 · answered Jun 27 '17 at 11:09

0

Try:

 ltsp-chroot -m 
 apt-get install linux-image-amd64/i386
 apt-get install linux-headers-amd64/i386
 ltsp-update-image 
 ltsp-update-kernels

answered Jun 27 '17 at 11:09

Hunter.S.Thompson

148
6

PXE-E11: ARP Timeout

4 Answers4