In the previous chapter, we covered a few patterns that allow servers to offload static data, or data that does not require request-time parsing, to AWS-backed storage. This allowed us to take advantage of storage that required very little maintenance operationally and, instead, focus on optimization of delivery through cache, content delivery networks, and so on.

These previous patterns are very rarely encountered in real-world exercises, as static data is generally rare itself. Real-world web applications are, by design, dynamic for the data is not set in stone, but changes over time through usage. From here, many issues arise from having multiple servers attempting to share data that cannot be cached easily, or at all.

One of the worst things that could happen from an end user's perspective would be to visit a web application and get conflicting data each time they change pages, or even refresh. If the underlying systems cannot properly share...