@george1421 Checking for the folder /sys/firmware/efi works perfectly so far on the systems we are testing.

If the folder exists we assign it the UEFI image, if it doesn’t exist it gets assigned the BIOS image.

@george1421 We use this particular kick off to “perform full host registration” of the device which assigns a base image and registers the host in fog as well as updating our inventory system with the device’s location, our barcode, etc.

Up until now we had everything configured for BIOS and have it hardcoded to assign our current base image. Now that I have two base images, was hoping to automate the assignment of the image to avoid in the field typos. I suppose we could just add in the option to chose UEFI or BIOS to the choose your own adventure during the initial full host registration of the device.

We have recently began building the support for UEFI within our environment. DHCP has been set to detect the vendor class of the device and boot to ipxe based on vendor class. A new UEFI base image has been tested and is working correctly. We have scripts to reorder the UEFI boot order for the machines.

What we are currently struggling with is how to get our post init script to kick off the UEFI or BIOS base image.

Currently we have the posit script to allow entry of barcode and location of the device. Then we pull device information and update Fog and Snipe-IT inventories . Lastly it assigns the base image.

What we are trying to figure out is how to have the script assign the correct BIOS or UEFI image to the machine. Is there a decent way to pull DCHP scope 67 information of the device to tell if ipxe.efi is used or not?