Managing many containers.
WHAT ORCHESTRATION DOES
Schedules containers across machines Restarts failed ones Scales the number running Routes traffic to them Manages configuration and secrets
WHEN IT IS WARRANTED
Many services Significant scale A team operating it
WHEN IT IS NOT
A single application A small team Anything one server handles comfortably
WHAT IT COSTS
Substantial complexity, and expertise to operate.
WHAT THAT MEANS PRACTICALLY
Orchestration introduces more failure modes than it removes, at small scale.
WHAT TO USE INSTEAD
Running containers directly on a server, with a process manager restarting them A managed platform handling this for you
WHAT TO ASK BEFORE ADOPTING IT
What problem are we solving that simpler options cannot?
WHAT TO BE HONEST ABOUT
Who will operate it, and what happens when they leave.
WHAT TO AVOID
Adopting it because it is standard practice elsewhere.
WHAT MOST APPLICATIONS NEVER OUTGROW
One well-configured server.
