Quick question here - according to the Kubernetes NodeJS doc, when I create a new deployment using

new k8s.apps.v1.Deployment(name: "some-name")

, shouldn't

some-name

become the resource name? Why does it become appended by a random string?

if you set the name property in the metadata block it will respect it but you have to check for uniqueness yourself after that

Autonaming allows us to improve the replacement semantics of kubernetes resources. So for example, if a config map's data is updated, pulumi can create a new config map, then update all pods to use the new version, and then only if that succeeds, delete the old one.