Skip to main content
Starting from version 4.45, Cognigy.AI supports Amazon EFS CSI driver as the EFS external provisioner is deprecated.

​
Prerequisites

  • Kubernetes v1.21 - 1.24.
  • Kubectl utility is installed locally.
  • Helm v3.8+ is installed on the client host.
  • Amazon EFS CSI driver is installed in the kubernetes cluster.
  • Cognigy.AI installation must be >= v4.45.
  • Snapshots/Backups of all EFS Persistent Volume Claims/Persistent Volumes (flow modules, functions) are must be ready before the migration starts.

​
Migration process

  1. Retain the flow-modules and functions Persistent volumes. 
    kubectl patch pv <flow-modules-pv-id> -p '{"spec":{"persistentVolumeReclaimPolicy":"Retain"}}'
kubectl patch pv <functions-pv-id> -p '{"spec":{"persistentVolumeReclaimPolicy":"Retain"}}'
  2. Enable the EFS CSI provisioner in the values.yaml file. 
    flowModules:
  persistence:
    aws:
      efs:
        enabled: true
        ## EFS File system ID
        ##
        id: ""
        efs_csi:
          enabled: true
functions:
  persistence:
    aws:
      efs:
        enabled: true
        ## EFS File system ID
        ##
        id: ""
        efs_csi:
          enabled: true
    For more details, check here
  3. Disable the currently running efs provisioner in the values.yaml file by deleting the following code block. 
    efs:
  flowModules:
    id: "<flow-module efs id>"
  functions:
    id: "<function efs id>"
  4. Delete the flow-modules and function StorageClass. 
    kubectl delete sc aws-efs-flow-modules aws-efs-functions
  5. Reapply the helm chart. 
    helm upgrade --namespace cognigy-ai cognigy-ai oci://cognigy.azurecr.io/helm/cognigy.ai --version HELM_CHART_VERSION --values YOUR_VALUES_FILE.yaml --create-namespace