ACADEMIA

Nothing demonstrates our dependence on technology more than when a server goes down and office productivity comes to a screeching halt. Author Karl Kopper's real-world response to this scenario is outlined in "The Linux Enterprise Cluster" (No Starch Press, April '05). The book describes how he put several inexpensive computers on a network, installed free software and made the individual machines operate as one big server called a cluster, thus saving his company time, money and resources. "The Linux Enterprise Cluster" goes through every step needed from start to finish to build a cluster of Linux computers capable of supporting an entire enterprise reliably and cost-effectively. Kopper explains that his book doesn't contain new or arcane technology. Rather, it contains all of the information and knowledge he gained while assembling an enterprise-class cluster from scratch for United Natural Foods, the largest publicly traded wholesale distributor to the natural and organic foods industry. Kopper did extensive legwork finding free or open source software and consulting countless websites, open source experts, help files, email lists and FAQs, as well as leveraging his own considerable experience supporting large servers and distributed computing environments. This is the only book to aggregate the most helpful information - based on actual experience - about enterprise-class Linux clusters in one reference manual. The actual cluster that served as the book's model went live in February 2003 and has experienced virtually no system downtime on a free operating system built cooperatively by volunteers on the Internet. Kopper sees "The Linux Enterprise Cluster" as a testament to the power of free software, online collaboration and the open source community. "After doing all the research and having success replacing a big monolithic server with a cluster, I was inspired to share my knowledge with other system administrators with a similar need," said Kopper. "It will be useful to a wide audience, since the cluster of individual computers can be virtually any size, depending on how much computing horsepower is required. The book shows that it's possible to build a very inexpensive and reliable business system for a small enterprise or large corporation." "The Linux Enterprise Cluster" is a valuable resource for system administrators who use Linux; system administrators interested in Linux but short on experience; students new to the information processing field; and anyone considering deploying Linux in an environment where low-cost computer reliability is important. In fact, any systems administrator charged with keeping mission-critical systems up and running will find useful techniques here; the book's second half describes high-availability techniques that can be used with or without a cluster. Also described in "The Linux Enterprise Cluster": * How to build a high-availability server using the Heartbeat package * How to configure a reliable printing systems in a Linux cluster environment * How to build a job scheduling system in Linux with no single point of failure ABOUT THE AUTHOR: Karl Kopper has worked for nearly 20 years with distributed computing environments on many platforms, including Linux, Windows, Macintosh, and UNIX. He has contributed to the development of numerous Linux and open source projects, including Heartbeat, Ganglia, Linux Virtual Server (LVS), Kernel and Zope. When not building high-availability clusters, he enjoys gardening. "The Linux Enterprise Cluster" by Karl Kopper April 2005, 536 pp., $49.95, ISBN 1593270364 Available at fine bookstores everywhere, from www.oreilly.com/nostarch, or directly from No Starch Press (www.nostarch.com, orders@nostarch.com, 800.420.7240).