Chapter Six · Part 2: The Toolchain
Orchestration
Chapter 5 packaged Pageturn into a container that runs the same anywhere. This chapter asks the next question: what happens when you need not one copy of it but hundreds, spread across many servers, kept alive day and night? Doing that by hand falls apart fast — so a piece of software takes over the job.
A single container is easy to start and watch. Real applications, though, don't run as a single copy. They run many copies at once — to serve more users than one copy could handle, and so that if one dies, the others keep going. The moment you have many copies across many servers, you inherit a pile of new chores: deciding what runs where, noticing when something crashes, replacing it, and swapping in new versions without going offline.
That whole job is called orchestration, and the software that does it is an orchestrator. These three topics build it up gently: first why one container is easy but hundreds is not, then what an orchestrator actually does for you, and finally how it rolls out a new version a few copies at a time so readers never see the site go down. Kubernetes is the best-known orchestrator, and this chapter is the on-ramp to it.