KubeCon + CloudNativeCon Europe 2025

So you've learned about Custom Resource Definition (CRD) design errors, you've designed your CRD to avoid common mistakes, and now you're ready to write the controller.

Turns out there's a lot of gotchas in that process as well!

This talk explores the common pitfalls that the ever-unlucky Charlie Don't, who always makes the worst decisions, runs into when implementing a controller.

The talk should be particularly useful for anyone writing reconciliation loops that use Kubernetes objects, whether they are CRDs or not. You can expect to come away from this talk having learned about common mistakes like: straining the apiserver with too many status updates, missing updates in complex systems of CRDs, and having scaling problems from not using caching correctly.

No knowledge of the previous talks is required, so come and have a chuckle at poor old Charlie Don't's bad luck while picking up some tips for yourself.