Kubernetes Multi-Cluster Lifecycle Management
I've been collaborating with some of the brilliant folks over at Google to help come up with a shared understanding of what safe upgrade processes look like in the world of Kubernetes -- with a focus on the idea of cluster management, eg. one level of abstraction up the chain for safe rollouts of entities within your clusters.
We're working towards a model of treating clusters entirely ephephemerally, in true "pets vs cattle" fashion. The first step is a five-part blog series hosted on Google Cloud's Medium space: a dive into terminology, thoughts on important considerations, and then some hands-on guides for a manual process using GKE's "Multi-Cluster Ingress for Anthos" product.
You can read all about it here!