Part 9 · 1 chapters · ~8 min

How a Company's Stack Grows

From a PaaS and one database to caches and queues, redundancy, containers and CI/CD, orchestration and tracing, gateways, secrets and flags, Kafka and meshes, and regulated multi-region operation, with the pain that justifies each addition.

12

Stage by stage

the staff habit
Every component added to a stack is something that can fail at 3 a.m. Write down the pain each one fixes and the signal that would let you remove it again. Stacks that grow by fashion become stacks nobody can operate.
HOW A STACK GROWS
what gets added at each stage, and the pain that triggers it
day 1PaaS, one app, onePostgres, Sentry
swipe the figure sideways, or tap expand for full screen
1/6
start small
A PaaS (Render, Fly, Heroku, App Platform), one app, one managed Postgres and error tracking. Most products should stay here far longer than engineers want.
PaaS + managed DB + error trackingstay here as long as it works