Working with resources

Networking (Services, Ingresses, Gateways, Network Policies)

Services

The Services list shows type, cluster IP, external IP, ports, and age. Click a row for detail.

When a Service appears in a Related Resources section (on Pod / Deployment / StatefulSet / Ingress / HTTPRoute detail views), its icon is tinted by endpoint health:

  • Green - every pod matching the selector is Running
  • Orange - some pods are Running, others aren’t
  • Red - no pods are Running, or the selector matches no pods
  • Default blue - selector-less services (headless / external Endpoints)

Service detail

The detail panel has three tabs:

  • Overview - service type, selectors, ports (with appProtocol shown when set), an advanced Networking section (traffic policies, source ranges, IP families) when those fields are present, a Pods section listing every pod whose labels match the service selector, and an Endpoint Health section for the live endpoint slices backing the service
  • Graph - visual dependency map
  • YAML - live manifest

Quick port-forward

For each declared service port, the detail view shows:

  • A port-forward arrow button - opens a tunnel to the service port via a ready backing pod
  • A clipboard button - copies the equivalent kubectl port-forward service/... command

See Port Forwarding for the full flow including unsafe-port handling and custom port persistence.

DNS lookup

The Service detail header has a DNS button that runs a DNS resolution from inside the cluster (via a temporary pod) so you can verify the service is reachable via its short name (my-svc) and fully-qualified name (my-svc.namespace.svc.cluster.local).

Ingresses

The Ingress list shows class, hosts, address, and age. Detail view structures the routing rules - host → path → backend service - and links to those services if they’re tracked.

GKE-specific annotations (networking.gke.io/managed-certificates, networking.gke.io/v1beta1.FrontendConfig) are surfaced as Related Resources so you can navigate to the ManagedCertificate or FrontendConfig CRD instance.

Gateways (Gateway API)

Available on GKE clusters and any cluster with the Gateway API CRDs installed. Detail view shows:

  • Gateway info - class, address, listeners (port + protocol + hostname)
  • Conditions - top-level Gateway status conditions (Accepted, Programmed, …) with reason + message + last-transition timestamp
  • Listener Conditions - per-listener status (one card per listener that has reported conditions) with attached-route counts
  • Related Resources - HTTPRoutes attached to the Gateway

HTTPRoutes

Routes through a Gateway. Detail view shows:

  • Routing - hostnames + parent Gateway
  • Conditions - per-parent acceptance status (one section per parentRef) with controller name and standard condition fields
  • Related Resources - parent Gateway + backend Services (icons tinted by endpoint health - see Related Resources note below)

Network Policies

Detail view shows policy types (Ingress / Egress / both) and the affected pods - Kubius resolves the pod selector against the live pod list so you can see which pods are governed.

When the selector is empty, the policy applies to every pod in the namespace and the detail view says so.

GCP-specific (GKE only)

GCPBackendPolicy

Full spec rendering: target ref, security policy, timeout, connection draining, logging, session affinity, custom headers. Status conditions are listed.

HealthCheckPolicy

Full spec rendering: target ref, check interval, timeout, healthy / unhealthy thresholds, logging, and per-protocol config (HTTP / HTTPS / HTTP/2 / gRPC / TCP).