kubernetes:<http://helm.sh/v3:Release|helm.sh/v3:Release> (ingress-nginx):
    error: could not get server version from Kubernetes: Get "https://<redacted ip>/version?timeout=32s": dial tcp <redacted ip>:443: i/o timeout

anyone know why the helm release resource seems to be making the above api call despite the cluster itself not being up yet? The cluster resource is currently in

depends_on

How did you construct the Kubernetes provider being used by the Helm resource? I’d think

depends_on

would handle it, but if you aren’t using an explicit provider you might need to do that.

annoyingly i can't find a good example programmatically generating the provider for GKE

Which language? The

kubernetes-gcp-{lang}

templates in https://github.com/pulumi/templates should have an example of generating a Kubeconfig for GKE.

oh i think this is exactly what i was looking for. I'm using python for devops, but we use go for prod code. ty!

Happy to help!