Working with resources

Pods

Pods are the central resource type in Kubius - most other resources (Deployments, Services, etc.) link back to the pods they manage or target.

Pod list

The Pods list shows every pod in the current namespace scope. Columns are sortable - click a header to sort, click again to reverse. By default, system-namespace pods are pushed to the bottom.

  • Status dot - green = Running, orange = Pending / ContainerCreating / not Ready, red = CrashLoopBackOff / Error, grey = Terminating
  • Restarts column highlights in red when above 5

Multi-select

Click a pod, then ⌘-click to add or remove pods from the selection, or shift-click to select a range. When two or more pods are selected the detail panel closes and a Combined Logs (N) pill appears in the toolbar - opens a multi-pod log viewer so you can tail every selected pod at once.

Row actions

Right-click any pod row for:

  • Open Resource - open in the detail panel
  • Open in New Window - detach into its own window
  • Open Terminal - exec into a running pod (disabled for non-Running pods)
  • Copy kubectl… submenu - each item shows the short command (e.g. kubectl get pod) but copies the full get/describe/logs/exec command with name + namespace
  • Pin / Unpin - pin to the sidebar for quick access
  • Delete Pod - graceful delete with confirmation (the watch removes the row once it’s gone). Force Delete isn’t offered from the right-click menu - use the pod detail Actions menu for that

The same right-click Delete X and Copy kubectl… are available on every resource list in the app, not just pods.

Pod detail

The detail panel has five tabs:

  • Overview - pod info, init containers, containers, conditions, related resources, labels, annotations (the Conditions block lives inside Overview rather than its own tab)
  • Events - Kubernetes Events for the pod, fetched via field selector and sorted newest first
  • Graph - visual dependency map (see Dependency Graph)
  • Logs - live tail with search and regex (see Logs)
  • YAML - live-loaded raw manifest, with a Diff pill against the last-applied configuration when present (see YAML)

Resource Usage section

The Overview tab shows live CPU and memory usage when the cluster has metrics-server installed. Toggle between gauges and history charts using the Mode button. Expand the charts with the pill next to the section title.

See Metrics for full details.

Container actions

Each container row in the Overview tab has:

  • Logs - jumps to the Logs tab pre-filtered to that container
  • Port indicators - click the quick-forward button next to a port to open a local tunnel, or the clipboard button to copy the kubectl port-forward command. See Port Forwarding.

Auto-discovered links to:

  • The pod’s ReplicaSet, Deployment, StatefulSet, DaemonSet, or Job owner
  • The Node the pod is scheduled on
  • Services that target the pod (label-selector match)
  • ConfigMaps and Secrets the pod mounts or references via env

Click any link to navigate. Right-click to open in a new window.

Pod actions

The Actions menu in the pod detail header offers:

  • Terminal - open an interactive shell in the pod (see Terminal)
  • Port Forward - open the full port-forward dialog
  • Restart Pod - deletes the pod; the controller (Deployment, StatefulSet, etc.) schedules a replacement
  • Evict Pod - evict via the policy/v1 Eviction API (same path kubectl drain uses). PodDisruptionBudgets are respected. The pod’s controller schedules a replacement - usually on a different node, but the same node is possible if no better fit exists. For DaemonSet pods, the kubelet immediately recreates the pod on the same node (DaemonSets pin to nodes by design - to move it, drain or cordon the node instead). Bare pods (no controller) won’t be replaced. PDB violations surface the offending budget’s message so you know what’s blocking
  • Delete Pod - permanent delete; if managed, a replacement is scheduled
  • Force Delete - gracePeriodSeconds=0. Use only for stuck pods - bypasses graceful shutdown

Pod Top

Available from the Pod list toolbar - opens a top-style live view of CPU and memory across every pod in the current namespace, sortable.