Helm install or upgrade release failed on Kubernetes cluster: the server could not find the requested resource or UPGRADE FAILED: no deployed releases

Tiller stores all releases as ConfigMaps in Tiller's namespace(kube-system in your case). Try to find broken release and delete it's ConfigMap using commands:

$ kubectl get cm --all-namespaces -l OWNER=TILLER
NAMESPACE     NAME               DATA   AGE
kube-system   nginx-ingress.v1   1      22h

$ kubectl delete cm  nginx-ingress.v1 -n kube-system

Next, delete all release objects (deployment,services,ingress, etc) manually and reinstall release using helm again.

If it didn't help, you may try to download newer release of Helm (v2.14.3 at the moment) and update/reinstall Tiller.

I had the same issue, but cleanup did not help also try the same helm chart on a brand new k8s cluster did not help.

So I found out that there was a missing apiVersion caused the problem. I found it out by doing a

helm install xyz --dry-run

copy the output to a new test.yaml file and use

kubectl apply test.yaml

there I see the error (apiVersion line was moved to a comment line)