fix(e2e): Correctly wait for a node pool stack for deletion

Author: mumoshu

e2e/run

@@ -335,6 +335,8 @@ nodepool_destroy() {
 
   if [ "$status" != "DELETE_IN_PROGRESS" ]; then
     ${KUBE_AWS_CMD} node-pools destroy --node-pool-name ${KUBE_AWS_NODE_POOL_NAME}
+    aws cloudformation wait stack-delete-complete \
+            --stack-name ${KUBE_AWS_NODE_POOL_STACK_NAME}
   else
     aws cloudformation wait stack-delete-complete \
         --stack-name ${KUBE_AWS_NODE_POOL_STACK_NAME}