blog

All field notes

Everything I've learned breaking (and fixing) infrastructure, written down before the memory fades.

6 min

Terraform state: the file that owns your weekend

Remote state, locking and why 'just re-run apply' is not a recovery strategy. Lessons from managing state across six teams.

  • Terraform
  • IaC
5 min

Kubernetes requests vs limits: stop guessing

CPU throttling mystery crashes, OOMKilled pods at 3 AM — most of it traces back to blindly copied resource values.

  • Kubernetes
  • SRE
7 min

How we put our AWS bill on a diet (-38%)

No magic, no downsizing prod into oblivion — just tagging discipline, rightsizing, spot orchestration and deleting the forgotten.

  • AWS
  • FinOps

looking for something older? ask me directly