Rebooting a Solaris 11 system to PXE/DHCP to completely rinstall the server

Question

Hi, I am working in an education application where we need to reinstall Solaris 11 x86 (or Solaris 8,9,10 or Linux,...) each week. The current approach has been to erase the MBR on the boot hard drive, and therefore let the host "fall-through" to the network boot device, which then initiates a PXE request to DHCP, and continue with Jumpstart, Kickstart or auto-install, as the server has configured at the time.
I cannot determine if finding the bootdrive and erasing the MBR is even feasible on Solaris 11 x86, but thought there might be a better way or at least a recommended way.
btw, The manipulation of GRUB menus to solve this presents a difficulty, as the client will not know in advance which type of reinstallation has been configured without probing the server somhow. This would have to be dynamically determined at each reinstall time. And, (not necessarily a bad thing) that approach would obsolete Jumpstart of Solaris 10 x86 03/05 which used real mode bootloader, not GRUB.
(SPARC is not a problem, as we just reboot -- net:dhcp and that hardware platform support is Solaris-only in any case).
Thanks for ideas.

Answer 1

I don't really have a better suggestion than clearing the MBR, it's the lack of ability to control the boot target configuration in the BIOS from the OS that gets in the way of doing more or less what you do on SPARC. You can determine the disk(s) in the root pool using zpool status rpool (assuming you're using the default root pool name rpool) and otherwise do the same things you currently do with Solaris 10.