Migrating a CTE install from Helm to Operator deployment

For CTE for Kubernetes v1.5.0 and subsequent versions, CTE for Kubernetes Operator is automatically installed by default.

Prior to v1.5.0, users had to manually migrate from a Helm-based deployment to Operator-based deployment. This is an optional feature for prior versions.

Upgrading to v1.5.0 from a prior version

If you are using a version prior to v1.5.0, for example, CTE for Kubernetes v1.3.0 deployed using Helm, then if you attempt to install CTE for Kubernetes v1.5.0, CTE for Kubernetes displays an error message stating that the existing deployment was completed using Helm. For that scenario, you have the following two options:

• Pass the --helm parameter to the deploy.sh script to upgrade to CTE for Kubernetes v1.5.0 using Helm charts. See |Upgrading||

• Migrate to operator based deployment using the following section:

Manually migrating to CTE for Kubernetes Operator

To switch to a Kubernetes operator deployment, you can manually migrate from Helm to Operator so that you can upgrade to CTE for Kubernetes v1.5.0 and subsequent versions:

1. Stop all of your applications that use CTE Volumes. Ensure that you stop all application pods using CTE-K8s and those actively accessing CTE volumes.

2. Revert the CTE for Kubernetes deployment, type:

   For CTE for Kubernetes v1.5 and subsequent versions

   ./deploy.sh --helm --remove

   For CTE for Kubernetes v1.4 and previous versions

   ./deploy.sh --remove

3. Deploy Kubernetes operator, type:

   ./deploy.sh --operator -t <image tag>
   

4. Redeploy the application pods that use CTE volumes.