You all have been helpful with networks in the past and I'm hoping you can assist me further.
Following this tutorial [https://www.raspberrypi.org/documentati ... utorial.md],
I was able to successfully boot a Raspberry Pi 3 client, using another Pi 3 as a DHCP server. But I need further help, and I need to know if what I'm trying to do is possible, and where to find explanations.
I want to achieve two things:
1) Network boot with an OS different than Raspbian whose only dependencies are kernel.img, bootcode.bin, fixup.dat (all four), and config.txt. The problem is, over network boot, the Pi 3 client boot ROM expects the following files to be present: https://github.com/raspberrypi/firmware ... aster/boot
Thus, TFTP fails every time since it could not find bcm2837.dtb and other missing files that I don't need. Any way to define my own network boot file dependencies instead? Or is there a way to trick the Pi into thinking I have these unnecessary files?
Once I get #1, I'd love to move the server to an actual Linux box.
2) Network boot a Pi 3 from a Linux computer. The dev who posted the tutorial I referenced claims (in the comments) to have this working by using a Linux machine as a server. I attempted this and added Option 43 (set to "Raspberry Pi Boot"), but I only got 4 requests, and no replies.
What exactly do I need to have in my Linux server configuration to get my Pi 3 to boot a custom OS image?