An attacker exhausts node memory by creating many workloads. Which combination helps reduce this risk?

1. Only obscuring namespace names
2. Allowing unlimited privileged Pods
3. Resource quotas and appropriate requests and limits
4. Disabling all Pod status reporting