Migrating Rancher to a New Cluster

If you are migrating Rancher to a new Kubernetes cluster, you don’t need to install Rancher on the new cluster first. If Rancher is restored to a new cluster with Rancher already installed, it can cause problems.

Prerequisites

These instructions assume you have created a backup and you have already installed a new Kubernetes cluster where Rancher will be deployed.

Migrating Rancher to a New Cluster - 图1caution

It is required to use the same hostname that was set as the server URL in the first cluster. If not done, downstream clusters will show as unavailable in the cluster management page of the UI, and you won’t be able to click inside the cluster or on the cluster’s Explore button.

Rancher version must be v2.5.0 and up

Rancher can be installed on any Kubernetes cluster, including hosted Kubernetes clusters such as Amazon EKS clusters. For help installing Kubernetes, refer to the documentation of the Kubernetes distribution. One of Rancher’s Kubernetes distributions may also be used:

1. Install the rancher-backup Helm chart

Install the rancher-backup chart, using a version in the 2.x.x major version range:

  1. Add the helm repository:

    1. helm repo add rancher-charts https://charts.rancher.io
    2. helm repo update

  2. Select and set CHART_VERSION variable with a 2.x.x rancher-backup release version:

    1. helm search repo --versions rancher-charts/rancher-backup
    2. CHART_VERSION=<2.x.x>

  3. Install the charts:

    1. helm install rancher-backup-crd rancher-charts/rancher-backup-crd -n cattle-resources-system --create-namespace --version $CHART_VERSION
    2. helm install rancher-backup rancher-charts/rancher-backup -n cattle-resources-system --version $CHART_VERSION

    Migrating Rancher to a New Cluster - 图2note

    The above assumes an environment with outbound connectivity to Docker Hub

    For an air-gapped environment, use the Helm value below to pull the backup-restore-operator image from your private registry when installing the rancher-backup Helm chart.

    1. --set image.repository $REGISTRY/rancher/backup-restore-operator

2. Restore from backup using a Restore custom resource

Migrating Rancher to a New Cluster - 图3Important:

Kubernetes v1.22, available as an experimental feature of v2.6.3, does not support restoring from backup files containing CRDs with the apiVersion apiextensions.k8s.io/v1beta1. In v1.22, the default resourceSet in the rancher-backup app is updated to collect only CRDs that use apiextensions.k8s.io/v1. There are currently two ways to work around this issue:

  1. Update the default resourceSet to collect the CRDs with the apiVersion v1.

  2. Update the default resourceSet and the client to use the new APIs internally, with apiextensions.k8s.io/v1 as the replacement.

    Migrating Rancher to a New Cluster - 图4note

    When making or restoring backups for v1.22, the Rancher version and the local cluster’s Kubernetes version should be the same. The Kubernetes version should be considered when restoring a backup since the supported apiVersion in the cluster and in the backup file could be different.

  3. When using S3 object storage as the backup source for a restore that requires credentials, create a Secret object in this cluster to add the S3 credentials. The secret data must have two keys - accessKey, and secretKey, that contain the S3 credentials.

    The secret can be created in any namespace, this example uses the default namespace.

    1. kubectl create secret generic s3-creds \
    2.   --from-literal=accessKey=<access key> \
    3.   --from-literal=secretKey=<secret key>

    Migrating Rancher to a New Cluster - 图5note

    Add your access key and secret key as values for accessKey and secretKey in the command above.

  4. Create a Restore object:

    During a migration, prune must be set to false. See the example below:

    1. # restore-migration.yaml
    2. apiVersion: resources.cattle.io/v1
    3. kind: Restore
    4. metadata:
    5.   name: restore-migration
    6. spec:
    7.   backupFilename: backup-b0450532-cee1-4aa1-a881-f5f48a007b1c-2020-09-15T07-27-09Z.tar.gz
    8.   prune: false
    9.   encryptionConfigSecretName: encryptionconfig
    10.   storageLocation:
    11.     s3:
    12.       credentialSecretName: s3-creds
    13.       credentialSecretNamespace: default
    14.       bucketName: backup-test
    15.       folder: ecm1
    16.       region: us-west-2
    17.       endpoint: s3.us-west-2.amazonaws.com

    Migrating Rancher to a New Cluster - 图6Important

    The field encryptionConfigSecretName should be used only if your backup was created with encryption enabled.

    If this applies, provide the name of the Secret object containing the encryption config file. If you only have the encryption config file, but don’t have the secret created in this cluster, use the following steps to create the secret:

    1. Create an encryption configuration file

    2. The command below uses a file named encryption-provider-config.yaml, with the --from-file flag. Run the below once the EncryptionConfiguration is saved in a file called encryption-provider-config.yaml:

      1. kubectl create secret generic encryptionconfig \
      2.   --from-file=./encryption-provider-config.yaml \
      3.   -n cattle-resources-system

  5. Apply the manifest, and monitor the Restore status:

    1. Apply the Restore object resource:

      1. kubectl apply -f restore-migration.yaml

    2. Watch the Restore status:

      1. kubectl get restore

    3. Watch the restoration logs:

      1. kubectl logs -n cattle-resources-system --tail 100 -f -l app.kubernetes.io/instance=rancher-backup

    4. Once the Restore resource has the status Completed, you can continue the cert-manager and Rancher installation.

3. Install cert-manager

Follow the steps to install cert-manager in the documentation about installing cert-manager on Kubernetes.

4. Bring up Rancher with Helm

Use the same version of Helm to install Rancher, that was used on the first cluster.

  1. helm install rancher rancher-latest/rancher \
  2.   --namespace cattle-system \
  3.   --set hostname=<same hostname as the server URL from the first Rancher server> \
  4.   --version x.y.z

Migrating Rancher to a New Cluster - 图7note

If the original Rancher environment is running, you can collect the current values with a kubeconfig for the original environment:

  1. helm get values rancher -n cattle-system -o yaml > rancher-values.yaml

These values can be reused using the rancher-values.yaml file. Be sure to switch the kubeconfig to the new Rancher environment.

  1. helm install rancher rancher-latest/rancher -n cattle-system -f rancher-values.yaml --version x.y.z