Support more fake Pods for nebula nodes

[kerthcet]

What would you like to be cleaned:

func virtualCapacity() corev1.ResourceList {
	return corev1.ResourceList{
		corev1.ResourceCPU:    resource.MustParse("1k"),
		corev1.ResourceMemory: resource.MustParse("10Ti"),
		corev1.ResourcePods:   resource.MustParse("1k"),
		nvidiaGPUResource:     resource.MustParse("1k"),
	}
}

This is the configuraiton for each fake node, but it's too small, for example, 1k for CPU can only available for 15 Pods with 64 CPUs, let's support one node with at least 1k workloads with 64C / 256Gi / 8GPUs per workload.

Why is this needed:

[kerthcet]

/help
/good-first-issue