01Kuberneteslearnk8s.io·18 min read
The initial migration is the easy part. Everyone's motivated, the greenfield work is exciting, and the team is paying attention. The problem is what happens 14 months later when the original champions have moved on and the platform is now maintained by people who weren't in the room.
Field noteRead this before you write the migration RFC. The "year two problem" is real — I've watched it happen at three companies. The section on knowledge transfer debt is worth the whole read.
02Terraformblog.gruntwork.io·12 min read
State drift. Remote state locking failures. The import block that silently succeeds but leaves your plan permanently broken. These aren't edge cases — they're the daily reality of running Terraform at scale, and the official docs treat them like footnotes.
Field noteBookmarked this after a 6-hour incident caused by a state lock that didn't release. The section on `terraform state mv` footguns alone is worth your 12 minutes. Send to whoever owns your IaC.
03Platform Engplatformengineering.org·9 min read
Backstage won't fix the fact that your platform team reports to a VP who reports to a different VP than your product teams. The tooling is a 10% problem. The other 90% is whether the people who build the platform have any incentive to talk to the people who use it.
Field noteThis reframes the whole IDP conversation. If your portal adoption is low, ask who owns the roadmap and whether they've done a user interview in the last quarter. The org chart is the product.
