Kubernetes Gateway API
Adopting Gateway API for cluster ingress: GatewayClass, Gateway and route resources, splitting responsibilities between platform and application teams, and choosing an implementation.
Kubernetes features
Kubernetes capabilities that need their own design and configuration work in a self-managed cluster.
Kubernetes network policies
Segmenting cluster traffic with NetworkPolicy: choosing a network plugin that enforces it, default deny patterns, namespace and pod selectors, egress control and testing policies.
Kubernetes audit logging
Configuring Kubernetes API audit logging: writing an audit policy, choosing levels per resource, log and webhook backends, retention flags and shipping audit events to a SIEM.
Kubernetes autoscaling
Scaling workloads and nodes in a self-managed cluster: HorizontalPodAutoscaler, the Vertical Pod Autoscaler add-on, in-place pod resizing, KEDA for event-driven scaling and node autoscaling.
Kubernetes GPU scheduling with dynamic resource allocation
Running GPU and accelerator workloads with Dynamic Resource Allocation: DeviceClass and ResourceClaim design, device sharing, how it differs from device plugins and its preemption limit.
Kubernetes Windows nodes
Running Windows Server workloads in Kubernetes: supported Windows versions and runtimes, Linux control plane, HostProcess containers instead of privileged pods, and pod fields Windows does not support.
Kubernetes validating admission policy
Enforcing cluster rules with CEL-based ValidatingAdmissionPolicy instead of webhooks: policies, bindings, parameters and rolling out with Warn and Audit before Deny.
