Fabrizio Pandini

Kubernetes automation relies on Prow, our CI/CD system designed for massive scale.
We are looking for new contributors to help maintain this critical component of our project.
Learn more on how you can contribute: docs.prow.k8s.io/docs/getting...

“Minimize workload disruption” should be your north star, but while it is ok to avoid unnecessary disruption, if you start treating your K8s nodes as a mainframe you will have bigger problems down the line.

“in-place updates are most useful for making changes that don't otherwise require a node drain or pod restart; for example: changing user credentials for the Machine. On the other hand, when the workload will be disrupted anyway, just do a rollout” ☝️☝️

“This is not immutable rollouts vs in-place updates; Cluster API considers both valid options and selects the most appropriate mechanism for a given change”! This is Big! kubernetes.io/blog/2026/01...

A "demystified" Gartner hype cycle graph including points like "trough of zero real adoption" and "slope of enterprise marketers jumping on the bandwagon too late"

I'm going to business hell for this one

First snowboarding lesson today! It was fun, but the next time I think about never stopping to learn please remind me about bingo

Cloud native mad pills of wisdom series. Episode #2

“For my cloud native apps I always use PDB with max unavailable equal to zero, you know better safe than sorry 🔒”

Cloud native mad pills of wisdom series. Episode #1

“We are pretty sophisticated with our cloud native apps. What is this PDB thingy you were talking about before? 👀”

This is brilliant! Well done!