remove extra check for scaling from zero

since the CanScaleFromZero function checks to see if both CPU and
memory are present, there is no need to check a second time. This also
adds some documentation to the CanScaleFromZero function to make it
clearer what is happening.

cluster-autoscaler/cloudprovider/clusterapi/clusterapi_nodegroup.go

@@ -243,12 +243,6 @@ func (ng *nodegroup) TemplateNodeInfo() (*schedulerframework.NodeInfo, error) {
 		return nil, err
 	}
 
-	_, cpuPresent := capacity[corev1.ResourceCPU]
-	_, memPresent := capacity[corev1.ResourceMemory]
-	if !cpuPresent || !memPresent {
-		return nil, cloudprovider.ErrNotImplemented
-	}
-
 	nodeName := fmt.Sprintf("%s-asg-%d", ng.scalableResource.Name(), rand.Int63())
 	node := corev1.Node{
 		ObjectMeta: metav1.ObjectMeta{

cluster-autoscaler/cloudprovider/clusterapi/clusterapi_unstructured.go

@@ -193,6 +193,8 @@ func (r unstructuredScalableResource) Taints() []apiv1.Taint {
 	return ret
 }
 
+// A node group can scale from zero if it can inform about the CPU and memory
+// capacity of the nodes within the group.
 func (r unstructuredScalableResource) CanScaleFromZero() bool {
 	capacity, err := r.InstanceCapacity()
 	if err != nil {