the implicit way: I bet it's very unusual to run terraform inside a cluster
2 ways where you need to reference ~/.kube: I don't want to run any commands before running terraform, moreover I need to involve myself with contexts in case I have more than 1 cluster
the exec plugin way: no example for GKE, and I thought it's not possible until I stumbled into this
I believe the way described by the last link is what should be recommended. But it's like not at all obvious where it's documented. At least it deserves a link form the hashicorp/helm documentation.
It's a documentation issue, so I believe the usual template doesn't apply here.
The methods described in the documentation:
terraform
inside a cluster~/.kube
: I don't want to run any commands before runningterraform
, moreover I need to involve myself with contexts in case I have more than 1 clusterI believe the way described by the last link is what should be recommended. But it's like not at all obvious where it's documented. At least it deserves a link form the
hashicorp/helm
documentation.