Resilience

I mentioned this concept a few times talking about network topologies, but it bears special mention here, because you'll be responsible for maintaining resiliency of your application on a network from both ends of a connection. If your application leverages any resources on a network, you'll have to account for the very likely possibility that there will be times at which those resources aren't actually available on that network. You'll need to write your code so that it still responds to its users in a reliable and stable manner in the event of such an outage.

Likewise, if your software is a dependency for other systems on your network, and it goes down, the best thing for you to do is have a strategy in place for rebounding from such an outage. There are a number of viable solutions in place for notifying your downstream consumers that you've recovered from an outage, each with their own strengths and costs in terms of resource usage or development time, and we'll discuss several of them later in this book. For now though, it is sufficient to consider this as you design your solutions and adjust those designs accordingly.