PXE Netboot from Windows 7 PC using TFTP32 E32 Timeout Issue

Posted on


I am trying to netboot a PC with no CD Drive and no USB Booting options (Available in BIOS but does not recognise USB, broken perhaps somehow).

Anyhow here is the log from TFTP32:

Rcvd DHCP Discover Msg for IP, Mac 00:1F:D0:8D:8B:09 [06/12 13:06:48.916]
DHCP: proposed address [06/12 13:06:48.917]
Rcvd DHCP Rqst Msg for IP, Mac 00:1F:D0:8D:8B:09 [06/12 13:06:51.113]
Previously allocated address acked [06/12 13:06:51.113]
Connection received from on port 2070 [06/12 13:06:51.125]
Read request for file <netbootpxelinux.0>. Mode octet [06/12 13:06:51.126]
Using local port 53708 [06/12 13:06:51.127]
Connection received from on port 2071 [06/12 13:06:53.125]
Read request for file <netbootpxelinux.0>. Mode octet [06/12 13:06:53.126]
Using local port 53709 [06/12 13:06:53.127]
Connection received from on port 2072 [06/12 13:06:57.136]
Read request for file <netbootpxelinux.0>. Mode octet [06/12 13:06:57.137]
Using local port 53710 [06/12 13:06:57.137]
Connection received from on port 2073 [06/12 13:07:03.122]
Read request for file <netbootpxelinux.0>. Mode octet [06/12 13:07:03.123]
Using local port 53711 [06/12 13:07:03.124]
TIMEOUT waiting for Ack block #1  [06/12 13:07:06.129]
TIMEOUT waiting for Ack block #1  [06/12 13:07:08.129]
Connection received from on port 2074 [06/12 13:07:11.086]
Read request for file <netbootpxelinux.0>. Mode octet [06/12 13:07:11.087]
Using local port 53717 [06/12 13:07:11.088]
TIMEOUT waiting for Ack block #1  [06/12 13:07:12.139]
TIMEOUT waiting for Ack block #1  [06/12 13:07:18.126]
TIMEOUT waiting for Ack block #1  [06/12 13:07:26.090]

Also when I ping the computer I get:

Pinging with 32 bytes of data:
Reply from Destination net unreachable.
Reply from Destination net unreachable.
Reply from Destination net unreachable.
Reply from Destination net unreachable.

Ping statistics for
    Packets: Sent = 4, Received = 4, Lost = 0 (0% loss),
PS C:Windowssystem32>

If anyone has any ideas as to how I might go about rectifying this it would be most appreciated.



Additional information that might be useful

  1. I am not using a crossover cable; however, the network card on my server computer is a gigabit one, I’m not sure if that means I still need a crossover cable, however. The two PCs are connected by only a fast Ethernet switch.

  2. The final objective of the netboot is to install Debian on the client computer. I am currently running an Apache server but I haven’t yet ascertained what the next steps will be. I am booting pxelinux.0 (not really sure yet what it actually is, I got it from the Debian tar.gz file I downloaded from the net following the guide here).


You definitely have a problem with the IP SubNets. With one machine on 10.20.21.x and the other on 192.168.2.x you are on completely different network types. You will likely want to speak with a network admin or someone who knows about networking. Both machines need to be within the same network and on the same SubNet unless you have opened your TFTP server’s switch to the other SubNet.

I ran into the same “timeout waiting for ack” problem a while back with tftpd32. What fixed the problem for me was temporarily disabling my firewall on the Windows computer. After it could communicate without the firewall interfering, my transfer went through. As mentioned in a different blog you may also want to make sure that you have configured everything correctly on the TFTP Server tab in the options (e.g. pxe compatibility) and make sure that the subnet for the DHCP server matches the subnet of the interface that you are plugging the computer into and that there is no other DHCP server on your network when you are working on this.

Turning off your firewall can be dangerous, so be careful and don’t forget to re-enable it!


Now that I’ve re-read that and thought about it more, it looks like it is definitely a problem with your ip addressing. From the ping output it looks like your windows computer has an IP of and the computer you want to deploy to gets an IP address of from tftp32. Since they are on different networks, with your described setup they have no way of talking with each other. In order for this to work, you would have to edit either the DHCP pool in tftp32 or change your computer’s IP to a static address in the correct network to match the DHCP pool.

I had the same problem, with “TIMEOUT waiting for Ack block #1” errors whenever a client attempted to get a file from my tftpd server.

I first tried enabling the TFTP UDP port 69 in the server firewall, but this did not help.
Eventually, the root cause of the problem proved to be on the client side, and not on the server side!
The reason is that TFTP sends the data back to the client using a dynamically allocated UDP port.
It is necessary to add a rule in the client firewall, allowing the client tftp application to receive data on any UDP port. That is C:WindowsSystem32tftp.exe on Windows clients.

Save your time and avoid DHCP issues on PXE environments; use your already in place DHCP infrastructure and set your PXE server (i.e. Serva) in proxyDHCP mode.
This way you do not have to deal with different subnetwork IP errors, like the one you are experiencing now, and you do not have to alter your current DHCP server configuration.

Leave a Reply

Your email address will not be published. Required fields are marked *