I've recently finished reading the DSC book but I'm still struggling to conceptualize the relationship between the configuration name in the LCM config and the node name in the configuration data file when the configuration name is something like Server2012R2. There is really only one example in the book about configuring time zones that touches on configuration names. Every other example I can find of configuration data files uses the actual server names for node names. Using a configuration name like Server2012R2 would everything in the configuration data file just fall under node name with a '*'?
How do other people approach this? Is it better to use the server name as the configuration name in the LCM and just list out all the servers separately in the configuration data?
Nothing's really "better." It's what works for you.
The ConfigurationName is simply what the LCM looks for when it's pulling a file. Using the computer name is just an easy way of tracking which file goes with which computer.
NodeName is part of a ConfigurationData block, which allows a single configurations script to output multiple MOFs, with each MOF "customized" for the node it's intended for.
But there's no "better." In the book, we tried to outline what you -can- do, and then you need to really think about your environment and how you want to manage it, your configuration scripts, and so on. What works best for you is super dependent on a zillion little "internal" things like who manages the scripts, what tools you might be using to track which MOF goes where, and so on.