Configure migrations to use service accounts

If you set up migrations and migration schedules using user accounts, you will encounter token expiration-related errors. To avoid these errors, Portworx, Inc. recommends setting up migration and migration schedules using service accounts.

In contrast to user accounts that expire after a specified interval of time has passed, service account tokens do not expire. Using service accounts ensures that you will not encounter token expiration-related errors. See the User accounts versus service accounts section of the Kubernetes documentation for more details about the differences between service accounts and user accounts.

Perform the following steps on the destination cluster to configure migrations to use service accounts.

Create a service account and a cluster role binding

1. Create a file called service-account-migration.yaml with the following content, adjusting the value of the metadata.namespace field to match the one you use to set up your migration:

apiVersion: v1
kind: ServiceAccount
  name: migration
  namespace: default

2. Apply the spec:

kubectl apply -f service-account-migration.yaml

3. Create a file called cluster-role-binding-migration.yaml with the following content, adjusting the subjects.namespace field to match with the namespace you wish to set up your migration:

kind: ClusterRoleBinding
  name: migration-clusterrolebinding
  kind: ClusterRole
  name: cluster-admin
- kind: ServiceAccount
  name: migration
  namespace: default
NOTE: The field is set to cluster-admin. For details about super-user access, see the User-facing roles section of the Kubernetes documentation

4. Apply the spec:

kubectl apply -f clusterrolebinding-migration.yaml

Create a kubeconfig file

1. Create a file called with the following content, adjusting the values of the SERVER and NAMESPACE variables to match your environment:


SERVICE_ACCOUNT_TOKEN_NAME=$(kubectl -n ${NAMESPACE} get serviceaccount ${SERVICE_ACCOUNT} -o jsonpath='{.secrets[].name}')
SERVICE_ACCOUNT_TOKEN=$(kubectl -n ${NAMESPACE} get secret ${SERVICE_ACCOUNT_TOKEN_NAME} -o "jsonpath={.data.token}" | base64 --decode)
SERVICE_ACCOUNT_CERTIFICATE=$(kubectl -n ${NAMESPACE} get secret ${SERVICE_ACCOUNT_TOKEN_NAME} -o "jsonpath={.data['ca\.crt']}")

cat <<END
apiVersion: v1
kind: Config
- name: default-cluster
    certificate-authority-data: ${SERVICE_ACCOUNT_CERTIFICATE}
    server: ${SERVER}
- name: default-context
    cluster: default-cluster
    namespace: ${NAMESPACE}
    user: ${SERVICE_ACCOUNT}
current-context: default-context

2. To create a kubeconfig file, enter the following commands:

chmod +x && ./ > ~/.kube/migration-config.conf

3. Set the value of the KUBECONFIG environment variable to point to the kubeconfig file you created in the previous step:

export KUBECONFIG=~/.kube/migration-config.conf

Generate a cluster pair

1. To generate a cluster pair using this service account, enter the following storkctl generate clusterpair command:

storkctl generate clusterpair mig-clusterpair --kubeconfig ~/.kube/migration-config.conf  > mig-clusterpair.yaml

2. Copy the mig-clusterpair.yaml file to your source cluster, modify the options section to match your environment and apply it. Depending on whether you want to configure Portworx for asynchronous or synchronous disaster recovery, follow the steps in one of the following pages:

You can use this cluster pair to create migrations or migration schedules. See the Synchronous DR for Kubernetes or the Asynchronous DR section for more details.

Last edited: Tuesday, May 4, 2021