Our team has "wasted" a fair bit of time researching and implementing all the bits and pieces to build pre-prod prod Kubernets environments, whereas Azure (and obviously) GCE's out of the box solutions make it so much easier.
Our team has "wasted" a fair bit of time researching and implementing all the bits and pieces to build pre-prod prod Kubernets environments, whereas Azure (and obviously) GCE's out of the box solutions make it so much easier.
Assuming the customer set that wants a supported orchestration tool, it would allow those AWS customers to more easily leave for Azure or GCE. Perhaps more importantly, it would allow existing customers to have credible on-prem dev, test, or disaster environments.
When you have the sort of market share AWS has, there's no real incentive to open those doors. I suspect this won't change until/if this market is more balanced.
I would love to have managed k8s on amazon. It wouldn't move me away because of all those surrounding technologies.
Minio, for example: https://github.com/minio/minio or Google's cloud storage. Both are compatible with the S3 api.
AWS is wonderful, but you always have to have tooling ready to not be held hostage by a vendor. And any vendor will hold you hostage once they have enough market dominance or enough of your business.
Have you ever considered "what if I didn't presuppose AWS?"