Multiple Gateways in PXE Boot / TFTP Timeout

[#1]

Hi

Until now, we had for each subnet an own PXE server.

From now on, there should only be one PXE server serving all subnets.

We changed all ip-helper addresses and tried to set the DHCP options (66/67) but we receive a TFTP-E32 Timeout. The ports for tftp is openend and I was able to get the boot file from an installed device.

I think the problem is the second gateway which appeared, which is acctually a layer 2 switch.

Do you have any ideas why this second entry appears?

Thank you for you help!

Melanie

[Marvin Rhoads]

Moved post to LAN Switching and Routing for correct categorization and better visibility....

Can you share the configuration of the interface with the DHCP helper configured on it as well as the DHCP Scope Options on your DHCP server?

[#3]

Hi Marvin

Thank you for your fast answer.

Interface  /VLAN with IP Helper:

interface GigabitEthernet0/X
switchport access vlan 32
switchport mode access
switchport port-security
no logging event link-status
no snmp trap link-status
spanning-tree portfast
end

interface Vlan32
ip address XX.XX.32.223 255.255.255.0
ip helper-address XX.XX.1.159
no ip redirects
no ip route-cache
end

I tried different DHCP options. In the attached printscreen you will see the option 66 and 67 with the path to the boot file (as it worked with a tftp client).

Best regards

Melanie

[Marvin Rhoads]

Hmm ok.

You don't by any chance also have an IOS DHCP pool on the switch do you?

I wonder if the "no ip redirects" on the SVI may be messing up PXE Boot.

I'm wondering why the gateway assigned for the subnet is not the same as the SVI with the ip helper address. there may be perfectly good reasons but it's a bit unusual at first glance.

Can you possibly span a switch port and capture the transactions to analyze the decodes?

[#5]

Thank you so much for your effort Marvin.

I have just removed the ip helper addresses and set it on the firewall and now.. it works!