Overall, Stack Exchange has banned best-practice questions. I agree with the accepted answer, which says that many questions about best practices tend to come down to:
I have no clue, plz give me a tutorial or code
The logical fallacy behind many "Best Practice" questions
Many people hope or assume that there must be some globally accepted way of building networks that would suddenly remove all doubts and fears they have about building something they're unfamiliar with.
Speaking very bluntly, this notion is mistaken; designs are based on budget, personnel, schedule, and technical requirements. It's rare for people to truly have the same combination of those things.
There are some axiomatic rules we all follow in IT (see The Practice of System and Network Administration for good examples). But even within those rules, there is still a lot of room for judgement calls. As RFC 1925 points out, "(10) One size never fits all"; in other words, what works for one group of people might not work for you.
So what do you do if you need to design something and you aren't sure how? Don't be afraid to ask questions about how to implement a design, assuming you document what you're doing. There are a lot of people on Network Engineering who can help. However, the take-away is that broad network design best-practices are just wishful thinking.
Most Good questions (not just "Best Practices") are situational:
At this point, I don't think we should get hung up on the term "best practices" itself as an automatic close trigger. It's possible to ask a good best practice question, if you include enough information about your situation. Personally, if people use the word best-practice in their question, I think they must include:
- A statement summarizing the business need: Networks cost money, and money comes from supporting a business. If we understand something about the business traffic that the network carries, we get much closer to understanding the tradeoffs the OP might be facing
- A description of what alternatives the OP has considered and why the OP is stuck: If the OP hasn't done enough research to tell us at least one possible solution to some problem, I argue that the OP is not ready to crowd-source the question on Stack Exchange. Stack Overflow members frequently summarize this by asking "What have you already tried?". I think this is relevant to best practice questions because it gives us some idea of what the OP wants and doesn't want.
Best practice questions should include (where relevant & possible):
- Specific information about HW Models, Firmware levels
- Topology Diagram
- Device configuration excerpts
Bad "Best Practice" questions don't describe a specific situation:
Alternatively, if someone comes to the main NE site looking for overall "best practices for [insert-something-here]" and provides few other details / context around what they have already tried, then we definitely should put those questions on hold as "primarily opinion based" or "too broad".
Even though the question isn't suitable for the main NE site, best practice questions and recommendations are welcome in Network Engineering Chat.