e2e

README

End-to-end testing on AAD Pod Identity

Get Started

To run the E2E tests in a given Azure subscription, a running Kubernetes cluster created through AKS Engine or Azure Kubernetes Service (AKS) is required. A service principal with Contributor role under the node resource group is also required. To obtain the service principal credentials for AKS, you can refer to here. For AKS Engine, if you have an existing cluster, search for the servicePrincipalProfile field in apimodel.json under the deployment folder. Otherwise, refer to here. Last but not least, an Azure KeyVault is required to simulate the action of accessing Azure resources. You can run a script to help you provision all of the Azure resources required during the test run:

# login as a user
az login

export SUBSCRIPTION_ID="<SubscriptionID>"
export RESOURCE_GROUP="<NodeResourceGroup>"
export KEYVAULT_NAME="my-keyvault"
export KEYVAULT_SECRET_NAME="test-secret"

./test/e2e/setup-e2e.sh

The E2E test suite extracts runtime configurations through environment variables. Below is a list of environment variables to set before running the E2E test suite.

Variable	Description
SUBSCRIPTION_ID	The Azure subscription ID.
RESOURCE_GROUP	The resource group of your Azure Kubernetes cluster.
AZURE_CLIENT_ID	The client ID of your service principal.
AZURE_CLIENT_SECRET	The client secret of your service principal.
AZURE_TENANT_ID	The Azure tenant ID.
KEYVAULT_NAME	The Azure KeyVault name.
KEYVAULT_SECRET_NAME	The name of the secret stored in the Azure KeyVault.
KEYVAULT_SECRET_VERSION	The version of the secret stored in the Azure KeyVault.
MIC_VERSION	The MIC version.
NMI_VERSION	The NMI version.
IDENTITY_VALIDATOR_VERSION	The identity validator version
SYSTEM_MSI_CLUSTER	Set to true if you are using an Azure cluster with system-assigned identity enabled.
ENABLE_SCALE_FEATURES	Set to true if you want to enable the scale features.
IMMUTABLE_IDENTITY_CLIENT_ID	The client ID of the immutable user-assigned identity created by running setup.sh.
NMI_MODE	The NMI mode (standard, managed).
BLOCK_INSTANCE_METADATA	Set to true if you want to run test cases related to block-instance-metadata feature.
SERVICE_PRINCIPAL_CLIENT_ID	The client ID of test service principal with Reader role to KeyVault resource group.
SERVICE_PRINCIPAL_CLIENT_SECRET	The client secret of test service principal with Reader role to KeyVault resource group.

Finally, to kick off a test run:

make e2e

Identity Validator

During the E2E test run, the image identityvalidator is deployed as a pod to the cluster to validate the identity assigned to it. The binary identityvalidator within the pod is essentially the compiled version of identityvalidator.go. If the binary execution returns an exit status of 0, it means that the pod identity and its binding are working properly. Otherwise, it means that the pod identity is not established. You can manually try out the identity validator by executing the following command:

cat <<EOF | kubectl apply -f -
apiVersion: "aadpodidentity.k8s.io/v1"
kind: AzureIdentity
metadata:
  name: keyvault-identity
spec:
  type: 0
  resourceID: <KeyvaultIdentityResourceID>
  clientID: <KeyvaultIdentityClientID>
EOF

cat <<EOF | kubectl apply -f -
apiVersion: "aadpodidentity.k8s.io/v1"
kind: AzureIdentityBinding
metadata:
  name: keyvault-identity-binding
spec:
  azureIdentity: keyvault-identity
  selector: keyvault-identity
EOF

kubectl run identityvalidator --image=mcr.microsoft.com/oss/azure/aad-pod-identity/identityvalidator:v1.8.13 --labels=aadpodidbinding=keyvault-identity -- --sleep

kubectl exec identityvalidator -- identityvalidator \
                                  --subscription-id "$SUBSCRIPTION_ID" \
                                  --identity-client-id "$AZURE_CLIENT_ID" \
                                  --keyvault-name "$KEYVAULT_NAME" \
                                  --keyvault-secret-name "$KEYVAULT_SECRET_NAME" \
                                  --keyvault-secret-version "$KEYVAULT_SECRET_VERSION"

# Check the exit status
echo "$?"

Documentation

Index

• func Cleanup(input CleanupInput)
• type CleanupInput

Functions

func Cleanup

func Cleanup(input CleanupInput)

Cleanup cleans up the test environment after a test case is finished running.

Types

type CleanupInput

type CleanupInput struct {
	Namespace *corev1.Namespace
	Getter    framework.Getter
	Lister    framework.Lister
	Deleter   framework.Deleter
}

CleanupInput is the input for Cleanup.