I've been doing a lot of work and learning lately about increasing the security in my home lab, particularly within my Kubernetes cluster.
So I've been leaning all about NetworkPolicies. As of now I've added policies to most of my externally facing apps, and it makes me very happy.
Now in the unlikely event of a pod getting compromised it won't be able to make any lateral movement across my network since the pods now only have access to strictly what they need access to. Before you could scan my entirely lab network, and before I added VLAN isolation a while back my entire network.
Now the only way an attacker could escalate is to compromise the host somehow, which is also unlike since none of my pods run as root.