Kubernetes GPU sharing explained: MIG partitioning, time-slicing, and MPS trade-offs, the device plugin and DRA model, and how to raise GPU utilization without hurting isolation in 2026.
Kyverno vs OPA Gatekeeper for Kubernetes policy as code: admission control, language model, mutation, image verification, and how to choose in 2026 - with a decision matrix.
OpenTelemetry logs and a unified telemetry pipeline: the Collector, log-trace correlation, OTTL processing, tail sampling, and cost control patterns for 2026.
Kubernetes in-place pod resize (v1.33 GA) changes CPU/memory without restarts. How it works, VPA integration, resize policies, cost impact, and production gotchas.
A Dapr deep-dive for 2026: building blocks, the sidecar model, service invocation, state, pub/sub, workflows, and how Dapr abstracts infrastructure on Kubernetes.
A 2026 Kubernetes multi-tenancy decision record: soft vs hard isolation, namespaces, hierarchical namespaces, vCluster, Capsule, and Kata - choosing the right tenancy model for your platform.
A hands-on NVIDIA Jetson and K3s edge AI cluster tutorial: provision nodes, enable GPU scheduling, deploy a vision model, and run inference at the edge.