In technology solutions projects, the main focus often goes to the design and build of the solution, and once the project passes the user acceptance test (UAT) and reaches the go-live day, everyone will be satisfied and consider the job successfully completed. Although this success is true to an extent, let’s look at it from a more realistic perspective.
Let’s assume you have proposed a technology solution to your customer or organization that should be used for the next five years to provide a promised value with full ROI. Also, let’s assume the design, build, and testing of the solution will take around eight months. In practical terms, the design and build is nothing compared to the entire solution lifecycle, which is 60 months.
Therefore, the actual value of the solution occurs during the 60 months and not during the 8 months’ period. That means whatever nice and fancy design you propose and build for your customer will have “No value” until it’s proven during this period. In other words, if you architect a network with a lot of interesting cutting-edge technologies – whether your goal is to upsell a product or to show some differentiation with advanced designs – this will not actually bring value to your customer’s business until it’s in operation.
Once the solution is in operation, any incident that requires complex troubleshooting will fall under the category of operational costs; the time the operation team requires to fix an issue is a cost, and this time may involve a service outage. Also, any difficulty the operation team faces while responding to any new business requirements means additional cost (time, efforts, etc.).
That’s why simplicity is a key design principle, because it translates to less complex operations and manageability of the technology solution. Also, from an architect/designer mindset, you should always consider the principle “build today with tomorrow in mind.” This simply means that you should look at the bigger picture and ask yourself the following questions: What type of business you are dealing with? What are the short-term and long-term business directions/strategies in the market? This will help you to ensure, to a large extent, that what you are designing and building today will be flexible enough to support tomorrow’s requirements, such as scalability.
One good example here is the way Cisco Intelligent WAN was architected. Apart from the APIC-EM controller, which makes operation life extremely simple, let’s look at its network architecture at a high level.
If you are designing an enterprise WAN with different connectivity models without considering the IWAN approach, you can simply use GRE or DMVPN with IPSEC over the internet links while you use the native WAN transport and routing over the (MPLS L3VPN) provider. Technically, this design will work 100%.
However, if you look at the bigger picture, you will see that you have a large number of sites that might be using different WAN providers, and some remote sites may only have internet or a combination of both. In large networks, you will most likely be dealing with primary and DR data centers. With this approach, the number of different routing protocols you will be dealing with, routes redistribution, filtering, and sub-optimal routing possibilities, will be very high. This will lead to high operational complexity and it ultimately will translate into higher OPEX.
In contrast, with the IWAN approach there will be a single “overlaid” routed domain across the different underlay WAN transports. This will offer simpler routing operations and more flexible design to support valorous connectivity model.
In addition, from vendors or IT solution providers’ sales point of view, the operation phase is very critical and important to generate more revenue as well as to maintain the solution value. Therefore, commonly IT solution providers or vendors focus on account farming during this phase, to generate more sales opportunities as well as to help the customer through various optimizations and technology exploitation activities , some of which aim to overcome operational complexities such as introducing automation to transactional activities to reduce the services and applications provisioning time and to avoid the possibility of human errors.