5.5.2 Distributed Objects: CORBA and Java RMI

		boot floppy, it will still try to mount a local partition as the root file system. There is some black magic involved to make it boot using the NFS directory obtained via the BOOTP. The root device used by a kernel is stored in the kernel image and can be altered with the rdev program, usually located in /usr/sbin. You want the root device to be the NFS file system, but no device file exists for this purpose, so you have to create one. You can do this with the following command:

		mknod /dev/nfsroot b 0 255

		This creates a dummy block device with special major and minor device numbers that have special meaning to the Linux kernel. This interprets the device as an NFS root file system when set as a root device with:

		rdev zImage /dev/nfsroot

		Now that the kernel's root device is set to be mounted via NFS, you can write the kernel to a floppy with the dd command:

		dd if=zImage of=/dev/fd0 bs=512

		After creating your first clone disk, you should test it on a node and make sure everything works. After this test, you can duplicate the floppy and clone all of your nodes at once. This is a little inconvenient for systems with more than 16 nodes. Instead, you could make 16 floppies, and clone your nodes 16 at a time. Yet another option, if you have network interface cards with PROMs, is to not create a boot floppy. Instead, you can put the kernel in a directory accessible by TFTP and use BOOTP to boot the nodes with a cloning kernel. After your system is up and running, you don't have to mess with clone floppies if you don't want to. You can clone nodes that already have an active operating system, by installing a clone kernel and rebooting. This can even be done by logging into the machine remotely.

		6.4 Basic System Administration

		Simply getting a cluster up and running can be a challenging endeavor. Once you've done it for the first time, it won't seem so difficult anymore. For your first cluster, it will probably be a lot easier to skip the cloning process. You can go ahead and install identical software from scratch on all of your nodes, including your worldly node. After you get a feel for how you are using the system, you can fine tune it by removing nonessential software from internal nodes and setting up a node cloning system.