just read this interesting take about how platform teams focus way too much on the wrong thing. everyone loves bragging about their fancy progressive rollouts and automated pipelines, but the real issue is usually just validation. we spend all this effort making sure the code actually reaches production, yet we have no idea if it's even working once it gets there.
deploying broken code faster is just a high-speed disaster . instead of obsessing over deployment frequency, maybe we should be looking at how we verify the state of the system. does anyone else feel like we are
optimizing for speed neglecting basic health checks?
more here:
https://thenewstack.io/solving-the-validation-problem/