Kickstart Process for a Linux Build
Although you don’t normally need to know exactly what’s happening during the kickstart process, it’s always handy to have a little quick reference. It’s a favourite question in an interview for a Linux related role. This information should be enough to describe the process. You might want to find out more about initrd, PXE or DHCP. The trick, as always, is trying to remember it when asked. I can’t help you there, I know it and I can’t always remember it…
A brief summary of the kickstart process:
Systems PXE logic broadcasts the MAC address
If it’s not a static address DHCP is used during the Kickstart Process.
DHCP recognises the discovery request and extends an offer of network information needed for new machine to boot. This includes :
- IP Address
- Default Gateway information
- Subnet mask
- IP address of the tftp server (or http server) holding the bootloader
- Full pathname of the boot loader file of the tftp server
The machine then applies the network information given and initiates a session with the server ti request bootloader
Bootloader then searches for its config file (which kernel and options such as initrd should be executed)
Assuming SYSLINUX – it’s located in the pxelinux.cfg directory and name HEX equivalent of the servers IP address (Maybe the MAC address)
The machine accepts this info and uncompresses the initrd image and kernel. The kernel then boots and starts the kickstart installation with the options supplied by the bootloader configuration file, including the kickstart server information.
For those who don’t like reading, I guess I should have mentioned this at the start of the article…