Copy a resource field into a ConfigMap or Secret

Exporting a Spec or Status field using the FieldExport resource

ACK controllers are intended to manage your AWS infrastructure using Kubernetes custom resources. Their responsibilities end after managing the lifecycle of your AWS resource and do not extend into binding to applications running in the Kubernetes data plane. The ACK FieldExport custom resource was designed to bridge the gap between managing the control plane of your ACK resources and using the properties of those resources in your application.

The FieldExport resource configures an ACK controller to export any spec or status field from an ACK resource into a Kubernetes ConfigMap or Secret. These fields are automatically updated when any field value changes. You are then able to mount the ConfigMap or Secret onto your Kubernetes Pods as environment variables that can ingest those values.

FieldExport is included by default in every ACK controller installation and can be used to reference any field within the Spec or Status of any ACK resource.

Using a FieldExport

The Spec and Status fields of the FieldExport custom resource definition are available in the API reference. For this example, we will be creating an S3 Bucket and exporting the Status.Location field into a ConfigMap.

apiVersion: s3.services.k8s.aws/v1alpha1
kind: Bucket
metadata:
  name: application-user-data
spec:
  name: doc-example-bucket
---
apiVersion: v1
kind: ConfigMap
metadata:
  name: application-user-data-cm
data: {}
---
apiVersion: services.k8s.aws/v1alpha1
kind: FieldExport
metadata:
  name: export-user-data-bucket
spec:  
  to:
    name: application-user-data-cm # Matches the ConfigMap we created above
    kind: configmap
  from:
    path: ".status.location"
    resource:
      group: s3.services.k8s.aws
      kind: Bucket
      name: application-user-data

Applying this manifest to the cluster will:

  1. Create a new S3 bucket called doc-example-bucket
  2. Create a ConfigMap called application-user-data-cm
  3. Create a FieldExport called export-user-data-bucket that will export the .status.location path from the bucket into the ConfigMap

After the reconciler has created the bucket, the application-user-data-cm ConfigMap looks like the following:

apiVersion: v1
kind: ConfigMap
metadata:
  name: application-user-data-cm
  namespace: default
data:
  default.export-user-data-bucket: http://doc-example-bucket.s3.amazonaws.com/

The ConfigMap data contains a new key-value pair. The key is the namespace and name of the FieldExport that created it, and the value is the resolved value from the resourc. This value can then be included as an environment variable in a pod like so:

apiVersion: v1
kind: Pod
metadata:
  name: my-application
spec:
  containers:
  - name: field-export-demo-container
    image: k8s.gcr.io/busybox
    command: [ "/bin/sh", "-c", "env" ]
    env:
    - name: USER_DATA_BUCKET_LOCATION
      valueFrom:
        configMapKeyRef:
          name: application-user-data-cm # The ConfigMap that we created earlier
          key: "default.export-user-data-bucket"

Looking at the container logs, you can see the USER_DATA_BUCKET_LOCATION environment is set with the value from the ConfigMap:

USER_DATA_BUCKET_LOCATION=http://doc-example-bucket.s3.amazonaws.com/

For a more comprehensive example of using FieldExport with multiple fields, check out the Deploy PostgreSQL + MariaDB Instances Using the ACK RDS Controller tutorial

FieldExport RBAC permissions

The ACK controller will fetch the source path from the ACK resource assuming its service account has the RBAC permissions to read that type of resource. If a user has the privileges to create a FieldExport resource, it is possible that they can create one which fetches fields from a resource they do not have RBAC permissions to read directly. This could potentially expose that resource’s properties to the unprivileged user.

To mitigate this problem, the ACK controller will only export fields from resources that exist in the same namespace as the FieldExport resource requesting it.

Edit this page on GitHub