We all learn better when we can safely deploy and break things. Time and time again, we know that things usually do not work right out of the box, and we need to spend time and energy to learn and experiment. This is especially true with newer technology such as SDN. By experimenting over and over with the same technology, you also gain muscle memories from repeated tasks. Do you remember the first time you logged in to the Cisco IOS or NX-OS shell? Did that feel strange? How about now after you have logged in perhaps hundreds of times? Often we learn best by doing.

The points I have made in this section might seem basic or common sense, but these are points I hold true throughout the years. I have seen projects not succeeding because some of the points mentioned previously were not followed through.

Perhaps a minimum viable product was not built that would have uncovered some detectable fault, or there was too much variation in the network for the new technology to adapt to. Hopefully, by preparing your network with familiarization of tools and frameworks, standardization, building minimum viable product, and relentlessly experimenting, you can decrease the friction of SDN migration for your own network.
In the next section, we will look at the consideration points for a greenfield SDN deployment.