Upgrading from v0.16 to v1.0

The upgrade process for upgrading to v1.0 is very Kubernetes version specific. Please check the version of your cluster using kubectl version and follow the steps required for your version of Kubernetes.

Issue with older versions of kubectl

kubectl versions with patch versions lower than v1.18.8 v1.17.11 or v1.16.14 have issues updating from the v0.16 CRD files, due to a bug when handling deeply nested CRDs. This bug will make kubectl apply -f [...] hang.

This bug only happens during a re-apply of the v0.16 CRDs or upgrading from it. Upgrades from lower versions do not cause issues. If you have this issue please upgrade your kubectl to the latest patch release. Versions of kubectl of v1.15.x or below are not being supported anymore as these are unsupported by the Kubernetes community.


Helm users who use installCRDs=true MUST upgrade to Helm v3.3.1 or later before upgrading.

Upgrade instructions per Kubernetes version

Kubernetes 1.16 and above

These are the upgrade instructions to upgrade from cert-manager v0.14.0 or higher, please consult other upgrade guides first before upgrading to v1.0 if you run an older version of cert-manager.

No special requirements, you can follow the regular upgrade process.

Kubernetes 1.15.x

cert-manager now uses apiextensions.k8s.io/v1 to install CRDs inside Kubernetes. This got added in Kubernetes 1.16. Our legacy installation will still be using apiextensions.k8s.io/v1beta1. For this reason Kubernetes 1.15 users now need to install the legacy version of the cert-manager manifests. You can follow the instructions of "Kubernetes 1.14" below on how to upgrade to the legacy version of v1.0.

Note: The legacy version only supports a single CRD version. We advise you to consider upgrading to Kubernetes 1.16 or above for an easier migration.

Kubernetes 1.14 and below

These are the upgrade instructions to upgrade from cert-manager v0.11.0 or higher, please consult other upgrade guides first before upgrading to v1.0 if you run an older version of cert-manager.

Note: Due to the lack of support for conversion webhooks in your Kubernetes version this will not be an easy migration. We advise you to consider upgrading to Kubernetes 1.16 or higher before upgrading. Upgrading your Kubernetes cluster might be easier than upgrading cert-manager.

We have released our cert-manager.io/v1 API that replaces cert-manager.io/v1alpha2. Since the legacy version for Kubernetes 1.15 and below only supports one CRD version you have to transition all resources to cert-manager.io/v1.

This makes for a fairly significant breaking change for users, as all cert-manager resources will need to be updated to reflect these changes. Ingress annotations will stay the same, this means if you only use ingress-shim you do not have to convert these resources over but it is recommended. However you should convert the (Cluster)Issuers and delete the old CRD versions.

This upgrade MUST be performed in the following sequence of steps:

  1. Back up existing cert-manager resources. See the backup section.

  2. Uninstall cert-manager.

  3. Update the apiVersion on all your backed up resources from cert-manager.io/v1alpha2 to cert-manager.io/v1. See the converting section for that.

  4. Ensure the old cert-manager CRD resources have also been deleted: kubectl get crd | grep cert-manager.io

  5. Re-install cert-manager v1.0 from scratch according to the installation guide.

  6. Apply the backed up resources again.

You must be sure to properly backup, uninstall, re-install and restore your installation in order to ensure the upgrade is successful.

Backing up resources

You can backup the custom resources you or cert-manager created using the following kubectl command:

kubectl get -o yaml \
--all-namespaces \
issuer,clusterissuer,certificates,certificaterequests > cert-manager-backup.yaml

Note that this will not export private keys or secrets.

Converting resources

You can use our kubectl plugin to automatically convert your backup from v1alpha2 to v1 using the following command:

kubectl cert-manager convert --output-version cert-manager.io/v1 -f cert-manager-backup.yaml > cert-manager-v1.yaml

Tip: you can use kubectl apply --dry-run on a local/test cluster with cert-manager v1.0 installed to validate your conversion

Uninstall cert-manager

Next step is to uninstall cert-manager. This will cause a temporary halt to renewal of certificates but will not affect any TLS traffic.

How you do this depends on how you installed cert-manager.

Using Helm:

$ helm --namespace cert-manager delete cert-manager

Using kubectl:

kubectl delete -f https://github.com/cert-manager/cert-manager/releases/download/vX.Y.Z/cert-manager.yaml

Make sure you also delete the CRDs. This will delete all cert-manager resources, so make sure your backup is complete. You can do this manually by executing the following commands:

kubectl delete crd certificaterequests.cert-manager.io
kubectl delete crd certificates.cert-manager.io
kubectl delete crd challenges.acme.cert-manager.io
kubectl delete crd clusterissuers.cert-manager.io
kubectl delete crd issuers.cert-manager.io
kubectl delete crd orders.acme.cert-manager.io

For more info see the uninstall cert-manager guide.

Reinstall and restore

To install cert-manager again you can follow the normal installation guide.

Once it has been fully installed you can re-apply the converted resources:

kubectl apply -f cert-manager-v1.yaml

Congratulations you're now fully upgraded to cert-manager v1.0