Hey List,

I apologize for how broad this may be, but hopefully someone here can help me out. I need to make custom kickstart installers to be used in the field to install microservers (Intel NUCs, EFI boot) that don't have optical drives and may not have internet connections.

I can make a custom ISO just fine using the pretty common steps, specifically following what's here:

http://smorgasbork.com/component/content/article/35-linux/151-building-a-cus...

My custom kickstart is on the disk, and along with all my packages, and everything works great on a VM as a CD image.

The problem is I can't get this to install from USB. I can get it to boot from USB using the isohybrid tool, but the installer fails, can't find the squashfs image. I don't think this is the right way to do it.

I know the CentOS 7 (and I believe 6.5) images have a "special" EFI partition that I believe is why USB works with the base image. The problem is I can't really find any information on how this works or how to recreate it. When I run fdisk -l on the image (as a loopback device) I see that the main image is on a partition marked as none, and then there's an EFI partition with start and end points actually inside the first partition? I'm afraid I don't really know what this is doing, or how to go about recreating.

Can anyone explain this to me or point me at some documentation? Am I missing something obvious, like say the old Revisor tool that I should be using to make my life much easier?

Any help would be much appreciated,

-Matt