My lab network runs on Linux, and has for almost a decade now. I need to test and support a variety of technologies, and Linux is the natural choice for that kind of scenario, given that it usually has the broadest base of infrastructure-oriented applications, and the most bleeding-edge features.

This doesn't mean that I savor instability, however. Indeed, a big part of my work depends on having an environment that is stable enough to facilitate rapid deployment of test systems and services, and reproducible testing. But I also need to be able to push the bounds of specific network-wide technologies, and that requires an infrastructure that is somewhat malleable. Linux gets me the combination of reliability and features, but it's actually a pretty tough balance to strike, since many of the Linux distributions tend to be optimized for other kinds of environments.

For example, some of the production-quality Linux distributions like Red Hat Enterprise Linux and Novell's SUSE Linux Enterprise excel on the stability part, but they do so by sacrificing some cutting-edge features, and they are always behind their counterparts when it comes to new technology. On the other end of the spectrum are bare-metal distributions like Gentoo Linux and the classic Debian, both of which are great for experimenting with brand-new technologies, but they don't provide the kind of long-term stability that I need for my lab.

For the past few years, I've been running SUSE Linux 9.3, because I've found that it provides the best combination of stability and features for my needs. The base operating system is stable and well-known, and it's also fairly easy to incorporate new technology when needed. But SUSE 9.3 is getting long in the tooth at this point, and some of the newer technologies are starting to require newer kernel support models, so it's well past time to move on.

The problem for me is that I haven't been able to find anything better than SUSE 9.3 for use in my labs. Even the SUSE 10 line has always come up short when measured against the 9.3 release. For example, SUSE 10.0 introduced a new hardware-detection model which wasn't very reliable, with some hardware only being discovered on every second reboot. Meanwhile, SUSE 10.1 had serious problems with its new package management model, which made it extremely difficult to maintain a reliable system. In both cases, there was no compelling reason to upgrade away from SUSE 9.3.