Topology & nodes

KubeMeridian reads the Kubernetes API directly (read-only) to render live inventory — no metrics required.

Cluster Status

The control-plane components plus a grid of nodes. Each node shows Ready / role / pressure badges, CPU & memory requests vs allocatable, pod count, and a grid of pod squares coloured by status (running / pending / failed) with a tooltip.

Applications Overview

A hierarchical map: namespace → workload → pod, covering deployments, statefulsets, daemonsets, jobs, cronjobs and bare pods, with the services that select them. Expand/collapse state is remembered.

• Click a pod → its Loki logs open in a drawer (when a Loki datasource is linked — see Logs & traces).
• Each pod links to the bundled Pod dashboard.

Nodes Overview

Per-node resource bars (CPU/memory requests vs allocatable, summed from the pods on the node), pod counts, roles, kubelet/OS/runtime info, and a deep link to the Node dashboard.

Events

A live, filterable feed of Kubernetes events — see Events.