Security for Kubernetess
Kubernetes security covers who can do what in a cluster and what Pods may do. Role-based access control grants permissions through Roles and bindings, NetworkPolicies restrict Pod traffic, and admission or policy engines such as Kyverno and OPA Gatekeeper reject workloads that break rules.
Stage: Security and Troubleshooting · Part of the Kubernetes roadmap · Last reviewed 19 November 2026
What to learn in Security
- Write a Role and RoleBinding that grants read-only access to Pods in a namespace
- Use ServiceAccounts for workloads and avoid default account permissions
- Apply a default-deny NetworkPolicy and then allow specific traffic
- Set securityContext options such as runAsNonRoot and read-only root filesystems
- Explain how an admission policy can block privileged containers
Security interview questions
What is the difference between a Role and a ClusterRole?
A Role grants permissions on resources within a single namespace. A ClusterRole is cluster wide and can cover non-namespaced resources such as nodes, or be bound in many namespaces. A RoleBinding grants a ClusterRole inside one namespace, while a ClusterRoleBinding grants it everywhere.
Do NetworkPolicies work without any extra setup?
No. NetworkPolicy objects are only enforced if the cluster network plugin supports them, for example Calico or Cilium. Without such a plugin the objects are accepted but have no effect. Pods are also open to all traffic until a policy selects them.
Where Security fits in the Kubernetes roadmap
Security belongs to the Security and Troubleshooting stage (advanced level), which takes about 22 hours. Keep it safe and healthy.
Other units in this stage
Back to the full Kubernetes roadmap