PXE Magic: Flexible Network Booting with Menus
One of my favorite features of a PXE server is the addition of a Knoppix rescue disk. Now, whenever I need to recover a machine, I don't need to hunt around for a disk, I can just boot the server off the network.
First, get a Knoppix disk. I use a Knoppix 5.1.1 CD for this example, but I've been successful with much older Knoppix CDs. Mount the CD-ROM, and then go to the boot/isolinux directory on the CD. Copy the miniroot.gz and vmlinuz files to your /var/lib/tftpboot directory, except rename them something distinct, such as miniroot-knx5.1.1.gz and vmlinuz-knx5.1.1, respectively. Now, edit your pxelinux.cfg/default file, and add lines like the one I used above in my example:
label 1 kernel vmlinuz-knx5.1.1 append secure nfsdir=10.0.0.1:/mnt/knoppix/5.1.1 nodhcp ↪lang=us ramdisk_size=100000 init=/etc/init 2 ↪apm=power-off nomce vga=normal ↪initrd=miniroot-knx5.1.1.gz quiet BOOT_IMAGE=knoppix
Notice here that I labeled it 1, so if you already have a label with that name, you need to decide which of the two to rename. Also notice that this example references the renamed vmlinuz-knx5.1.1 and miniroot-knx5.1.1.gz files. If you named your files something else, be sure to change the names here as well. Because I am mostly dealing with servers, I added 2 after init=/etc/init on the append line, so it would boot into runlevel 2 (console-only mode). If you want to boot to a full graphical environment, remove 2 from the append line.
The final step might be the largest for you if you don't have an NFS server set up. For Knoppix to boot over the network, you have to have its CD contents shared on an NFS server. NFS server configuration is beyond the scope of this article, but in my example, I set up an NFS share on 10.0.0.1 at /mnt/knoppix/5.1.1. I then mounted my Knoppix CD and copied the full contents to that directory. Alternatively, you could mount a Knoppix CD or ISO directly to that directory. When the Knoppix kernel boots, it will then mount that NFS share and access the rest of the files it needs directly over the network.
Another nice addition to a PXE environment is the memtest86+ program. This program does a thorough scan of a system's RAM and reports any errors. These days, some distributions even install it by default and make it available during the boot process because it is so useful. Compared to Knoppix, it is very simple to add memtest86+ to your PXE server, because it runs from a single bootable file. First, install your distribution's memtest86+ package (most make it available), or otherwise download it from the memtest86+ site. Then, copy the program binary to /var/lib/tftpboot/memtest. Finally, add a new label to your pxelinux.cfg/default file:
label 3 kernel memtest
That's it. When you type 3 at the boot prompt, the memtest86+ program loads over the network and starts the scan.
There are a number of extra features beyond the ones I give here. For instance, a number of DOS boot floppy images, such as Peter Nordahl's NT Password and Registry Editor Boot Disk, can be added to a PXE environment. My own use of the pxelinux menu helps me streamline server kickstarts and makes it simple to kickstart many servers all at the same time. At boot time, I can not only indicate which OS to load, but also more specific options, such as the type of server (Web, database and so forth) to install, what hostname to use, and other very specific tweaks. Besides the benefit of no longer tracking down MAC addresses, you also can create a nice colorful user-friendly boot menu that can be documented, so it's simpler for new administrators to pick up. Finally, I've been able to customize Knoppix disks so that they do very specific things at boot, such as perform load tests or even set up a Webcam server—all from the network.
Syslinux PXE Page: syslinux.zytor.com/pxe.php
Red Hat's Kickstart Guide: www.redhat.com/docs/manuals/enterprise/RHEL-4-Manual/sysadmin-guide/ch-kickstart2.html
Kyle Rankin is a Senior Systems Administrator in the San Francisco Bay Area and the author of a number of books, including Knoppix Hacks and Ubuntu Hacks for O'Reilly Media. He is currently the president of the North Bay Linux Users' Group.
Kyle Rankin is a director of engineering operations in the San Francisco Bay Area, the author of a number of books including DevOps Troubleshooting and The Official Ubuntu Server Book, and is a columnist for Linux Journal.
Webinar: 8 Signs You’re Beyond Cron
11am CDT, April 29th
Join Linux Journal and Pat Cameron, Director of Automation Technology at HelpSystems, as they discuss the eight primary advantages of moving beyond cron job scheduling. In this webinar, you’ll learn about integrating cron with an enterprise scheduler.Join us!