input

func (Object) Unstructured ¶

func (u Object) Unstructured() *unstructured.Unstructured

func (Objects) Len ¶

func (u Objects) Len() int

func (Objects) Less ¶

func (u Objects) Less(i, j int) bool

func (Objects) Swap ¶

func (u Objects) Swap(i, j int)

func (*ResourceRepo) AllNamespacesPodMetrics ¶ added in v0.7.19

func (r *ResourceRepo) AllNamespacesPodMetrics() (Objects, error)

AllNamespacesPodMetrics attempts a single cluster-wide list of PodMetrics (across every namespace), so that a render touching many namespaces/nodes only needs one metrics.k8s.io request instead of one per namespace. The attempt, and its outcome, is cached for the whole render (including across multiple nodes in the same query) so callers never retry it. If the cluster-wide list is denied (e.g. RBAC only grants namespace-scoped access), a warning is printed once and callers should fall back to per-namespace fetches, which may in turn be incomplete for namespaces the user also can't access.

func (*ResourceRepo) CLIQueryResults ¶

func (r *ResourceRepo) CLIQueryResults(args []string) *resource.Result

func (*ResourceRepo) DynamicObject ¶

func (r *ResourceRepo) DynamicObject(gvr schema.GroupVersionResource, namespace string, name string) (Object, error)

func (*ResourceRepo) EndpointSlices ¶ added in v0.7.15

func (r *ResourceRepo) EndpointSlices(namespace string) (*discoveryv1.EndpointSliceList, error)

func (*ResourceRepo) FirstObject ¶

func (r *ResourceRepo) FirstObject(namespace string, args []string, labelSelector string) (Object, error)

func (*ResourceRepo) GVRFor ¶

func (r *ResourceRepo) GVRFor(resourceOrKindArg string) (schema.GroupVersionResource, error)

func (*ResourceRepo) Ingresses ¶

func (r *ResourceRepo) Ingresses(namespace string) (*netv1.IngressList, error)

func (*ResourceRepo) KubeGetNodeStatsSummary ¶

func (r *ResourceRepo) KubeGetNodeStatsSummary(nodeName string) (Object, error)

KubeGetNodeStatsSummary returns this structure > kubectl get --raw /api/v1/nodes/{nodeName}/proxy/stats/summary The endpoint that this function uses will be disabled soon: https://github.com/kubernetes/kubernetes/issues/68522

func (*ResourceRepo) MetricsUnavailableReason ¶ added in v0.7.19

func (r *ResourceRepo) MetricsUnavailableReason() string

MetricsUnavailableReason reports why the metrics.k8s.io API (served by metrics-server) can't be used, or "" if it's healthy. It checks the APIService's Available condition directly -- the same signal `kubectl top` relies on to report "Metrics API not available" -- rather than a RESTMapper/discovery-based check, since discovery listings may still show a stale/cached group version while the backing service is down. This distinguishes two different problems a caller should word differently:

• the APIService object doesn't exist at all: metrics-server was never installed.
• the APIService exists but its Available condition is False: metrics-server is installed but currently unhealthy (e.g. crashed, unreachable); the condition's own message/reason is surfaced since it usually names the concrete problem (e.g. "no endpoints available").

The result is cached since Pod/Node rendering checks this repeatedly.

func (*ResourceRepo) NonTerminatedPodsOnTheNode ¶

func (r *ResourceRepo) NonTerminatedPodsOnTheNode(nodeName string) (Objects, error)

func (*ResourceRepo) ObjectEvents ¶

func (r *ResourceRepo) ObjectEvents(u *unstructured.Unstructured) (*corev1.EventList, error)

func (*ResourceRepo) Objects ¶

func (r *ResourceRepo) Objects(namespace string, args []string, labelSelector string) (Objects, error)

func (*ResourceRepo) Owners ¶

func (r *ResourceRepo) Owners(obj Object) (owners Objects, orphans []metav1.OwnerReference, err error)

Owners resolves the ownerReferences of obj to the objects they point at. References whose owner no longer exists are reported back as orphans, so callers can warn about them, rather than being silently dropped. Owners that can't be resolved for other reasons (e.g. missing permissions, unknown kind) are silently skipped, since that's not evidence the owner is gone.

func (*ResourceRepo) PodContainerLogs ¶ added in v0.7.19

func (r *ResourceRepo) PodContainerLogs(namespace, podName, containerName string, previous bool, tailLines int64) (string, error)

PodContainerLogs returns up to tailLines lines of log output for the named container in the named pod. When previous is true it fetches logs from the container's previous (terminated) instance, equivalent to `kubectl logs --previous`.

func (*ResourceRepo) Service ¶

func (r *ResourceRepo) Service(namespace, name string) (*corev1.Service, error)

func (*ResourceRepo) Services ¶

func (r *ResourceRepo) Services(namespace string) (*corev1.ServiceList, error)