How to determine what a host boots

Assuming that a client host is configured (in its BIOS) to boot from the NIC, the first point at which we can influence from the server how the client boots is in the DHCP table. A hosts place in the DHCP tables typically changes when it moves from one subnet to another, or when it changes from Windows-only to Linux-only. Generally, in order to use PXE for booting, it must be in a scope that offers the settings we described in Conference under Docker.

The second point at which we can influence the client from the server is at the PXE configuration file. At the IWI, this is done with the sethostip command, which takes two parameters: first name of a file already present in /tftpboot/pxelinux.cfg, then the hostname. See Conference under Docker.

Example 2.  Example of sethostip usage

The command sethostip -t bootpxe-sarge.sda iwi101 would make the host iwi101 boot as specified in the pxe configuration file /tftpboot/pxelinux.cfg/bootpxe-sarge.sda.


To be continued with exports, hosts.client, /etc/bundle, gen_domainprof, overwrites, addons, NEWDIST, MAKEPARENTDIST etc.etc.