You make some architectural decision, which makes lots of sense at the time. Over time, the environment changes a little here, and a little there, so that your decisions are no longer optimal. You then get promoted, and your new replacement is evaluating your work. If you've been on either side of this situation, you know that the replacement will evaluate the design based on the environment TODAY, and not the environment when it was designed. If you haven't documented why you made the decisions you did, it will look like you are a very poor decision maker.
When you document the "why", be sure to answer the following questions:
What decision did you make and when did you make it?
What constraints played a factor?
What assumptions played a factor?
What alternatives did you consider, and why were they not selected?
Keeping the answers to these questions in a repository where you can quickly look them up and understand them will not only help future generations of network administrators avoid landmines, but over time, it will also help you understand how you make decisions, and how your choices turned out, and thus, help you make better decisions in your own future.
