Fly.io Made Crystal Clear · chapter 3: Your First Deploy
The quickstart sequence, and what exists in the world after each step
2026-11-02
Nothing here is a leap of faith; every step is checkable before you move to the next.
Below: the paragraph from the book that builds this idea, then the diagram itself (Figure 3.2), and a recap. About a minute of reading.
While it runs, the output is answering three questions in order: did the build succeed, did each new Machine start, and did each new Machine pass its checks. A smoke check watches every Machine for about 10 seconds after it enters the started state, just to confirm it stays running rather than crashing immediately. Once a Machine is up, Fly.io's proxy runs the health checks defined in your config against it, and a Machine only joins the pool serving real traffic after it passes them. Fail, and the deploy either halts (for a rolling or canary deploy) or simply never cuts traffic to the new Machines (for bluegreen). Nothing routes to something unhealthy by accident.
Recap
- The idea: Nothing here is a leap of faith; every step is checkable before you move to the next.
- The picture: Figure 3.2, from chapter 3 ("Your First Deploy") of Fly.io Made Crystal Clear.
- Go deeper: the chapter builds this step by step, with recipes and sources at the end.
This diagram is one of many in Fly.io Made Crystal Clear.
Every chapter opens with the gist, draws the hard ideas, and ends with recipes and sources.
Get the book
