Cert-manager: Install on new kubeadm 1.13.0 cluster fails with "certificates.certmanager.k8s.io" already exists

[dkirrane]

Describe the bug:
Helm install fails on a newly created kubeadm-1.13.0 cluster:
Error: release cert-manager failed: customresourcedefinitions.apiextensions.k8s.io "certificates.certmanager.k8s.io" already exists

Expected behaviour:
Install works

Steps to reproduce the bug:
Create a new kubeadm cluster

helm repo update
helm install --name cert-manager  --namespace kube-system  stable/cert-manager

Anything else we need to know?:

Environment details::

• Kubernetes version: 1.13.0
• Cloud-provider/provisioner (e.g. GKE, kops AWS, etc): kubeadm-1.13.0
• cert-manager version (e.g. v0.4.0): stable/cert-manager v0.5.2
• Install method (e.g. helm or static manifests): helm v2.12.0

/kind bug

[Ryosuke-Sasaki]

My environment is GKE, but a similar phenomenon happened to me as well. Although it is not a fundamental solution, I installed version 0.4.1 and I successfully installed it.

[yfxie]

@Ryosuke-Sasaki 👍 issue solved by

 helm install --name cert-manager --namespace kube-system --version v0.4.1 stable/cert-manager

[michaelsteven]

Ran into these symptoms deploying on a new/clean 1.11.5 aks cluster today and the workarounds in https://github.com/jetstack/cert-manager/issues/870 didn't work. I was able to work-around and deploy the latest version by first doing a helm install --set createCustomResource=false, then following that up with a helm upgrade without setting that variable.

[clehene]

I got the same error and did a 0.4.1 install followed by an upgrade

[alexellis]

I'm also running into the same error with v0.5.2

[ahmetb]

This is likely the same thing as #1152 (uninstallation doesn't clean up the CRDs).

[marc-barry]

I ran into this exact issue on a DigitalOcean cluster with Kubernetes 1.13.1. I ultimately used https://github.com/jetstack/cert-manager/issues/870#issuecomment-446680412 to work around the issue.

[hjet]

Ran into this same issue while writing How to Set Up an Nginx Ingress with Cert-Manager on DigitalOcean Kubernetes. For now have included the workaround in https://github.com/jetstack/cert-manager/issues/870#issuecomment-446680412 and linked back to this issue, but will follow and update once resolved

[axeluhl]

Same thing here. I tried

kubectl delete customresourcedefinitions.apiextensions.k8s.io/certificates.certmanager.k8s.io; kubectl delete customresourcedefinitions.apiextensions.k8s.io/clusterissuers.certmanager.k8s.io; kubectl delete customresourcedefinitions.apiextensions.k8s.io/issuers.certmanager.k8s.io

and then a clean
helm install stable/cert-manager --version v0.5.2
but still got
Error: release ... failed: customresourcedefinitions.apiextensions.k8s.io "certificates.certmanager.k8s.io" already exists

helm install stable/cert-manager --version v0.4.1

worked without problems, and so did

helm upgrade existing-stingray stable/cert-manager

This is seriously broken...

[hjet]

This is resolved by upgrading Helm to v2.12.1 per https://github.com/jetstack/cert-manager/issues/870#issuecomment-450643669

[munnerz]

We've also moved to install the CRDs as a complete separate install step as of the upcoming v0.6 release, in order to avoid this and a few other issues with CRD installation and Helm.

I'm going to close this issue now, as I don't think there's much we can do as this is a Helm bug (that appears to be resolved :tada:). Feel free to reopen if you disagree, using /reopen.

Thanks for the report!

/close

[jetstack-bot]

@munnerz: Closing this issue.

In response to this:

We've also moved to install the CRDs as a complete separate install step as of the upcoming v0.6 release, in order to avoid this and a few other issues with CRD installation and Helm.

I'm going to close this issue now, as I don't think there's much we can do as this is a Helm bug (that appears to be resolved :tada:). Feel free to reopen if you disagree, using /reopen.

Thanks for the report!

/close

Instructions for interacting with me using PR comments are available here. If you have questions or suggestions related to my behavior, please file an issue against the kubernetes/test-infra repository.