Unix Articles & FAQ

First of all, write a kickstart file. This can be generated using the “Kickstart Configurator” application (the package is system-config-kickstart*.rpm on disc5), or by hand.

Example:

install
graphical
url --url http://10.211.55.4/rhel5-install/
bootloader --location mbr --driveorder hda --append "rhgb quiet"
lang en_US.UTF-8
keyboard us
clearpart --all
autopart
rootpw --iscrypted 7o0ppLxgDVmGc
selinux --enforcing
firewall --enabled --ssh --http --smtp
key --skip
timezone --utc US/Eastern
reboot
%packages --ignoremissing
@ X Window System

Fire up a virtual machine and boot from disc1, at the command prompt type:

linux ks=http://ip-address/path-to-kickstart-file

then hit return. Obviously, this assumes that you have the above file available on an HTTP server somewhere, doesn’t matter what it’s called, I called mine kickstart.ks.

Note the third line in the Kickstart file. How the hell do you set that up? Read on.

What you have to do is mount via loopback each ISO image under the rhel5-install directory, which should be available over HTTP.

For each disc, do this:
mount -o loop /path-to-isos/rhel5-disc1.iso /path-to-http-server/public/rhel5-install/disc1

For some annoying reason, OS X doesn’t give you the option of a loopback device, so I did this on a seperate Linux VM using a very basic HTTP server called abyssws. Mounting via loopback, by the way, just helps save space as you’re mounting each disc “in place”, rather than copying over files and losing more valuable time. My ISO images were actually shared from OS X to Linux, where they were then mounted using the method above.