Digital infrastructure using Kubernetes previously used a “bulldozer” approach, rebuilding entire sections to fix minor technical issues like security patches. New “in-place” update developments now allow for “surgical repair,” fixing specific parts without affecting the entire system.
Kubernetes infrastructure management is shifting from “immutable infrastructure” using a “cordon and drain” process for repairs to a “mutable infrastructure” approach with “in-place updates,” allowing for “chained upgrades” and reducing “configuration drift”. This change affects both the “Control Plane” and the “Data Plane” elements of a cluster.
> “configuration drift”
Hilarious. i.e. we don’t want to touch the stuff that seems to be working ;-)