I've seen some argument on how we should not run a database on Kubernetes in production, for various reasons, e.g, https://www.youtube.com/watch?v=Nosa5-xcATw&feature=youtu.be&t=1080. And how PetSet / StatefulSet is still on beta as of 1.5.
I wonder if the same holds true for OpenShift Origin. It is built on top of Kubernetes, but maybe, it has some "bells and whistles" for running database on production?
From what I see in https://github.com/openshift/origin:
Stateful Sets are still being actively developed and no backwards compatibility is guaranteed until 1.5 is released. Starting in 1.5, Stateful Sets will be enabled by default and some backwards compatibility will be guaranteed.
Statefulsets were recently available in Openshift (April 2017), and from what I see here: https://blog.openshift.com/kubernetes-state-storage/ in principle I do not see anything that differentiates using StatefulSets in Kubernetes and Openshift.