Skip to main content

Ingress

The Ingress tab shows the cluster's ingress controller and load balancer. Change gateway (top right) picks which ingress load balancer the console monitors when a cluster runs more than one controller.

Cluster ingress

The right rail describes the controller: Controller service (name, namespace, type, external / cluster IP) and the exposed Ports (each port → targetPort with its nodePort). The Control plane view lists backend servers with per-server Status, Sessions and In / Out traffic.

When the ingress is exposed through a LoadBalancer, extra tabs appear:

  • Statistics — controller status, backend-server health and live traffic.
  • Monitoring — utilisation charts for the load balancer.
  • Port forwarding — expose raw TCP/UDP ports straight to worker nodes (below).

The Load balancer rail card also offers Resize (a bigger/smaller flavor):

note

Resizing the ingress load balancer briefly powers it off — ingress traffic pauses during the swap. Resize during a maintenance window.

Port forwarding​

On a LoadBalancer ingress, Port forwarding exposes a public TCP/UDP port directly to selected worker nodes — an L4 path separate from the HTTP ingress rules. Quick forward port opens a dialog to set the Protocol (TCP/UDP), Public port and Node port, and to pick which worker nodes receive the traffic.

Existing forwards are listed as cards (protocol, port, backend node count); expand one to see its backend nodes, and the delete icon removes it.

Removing a forward stops traffic

Deleting a forward immediately stops traffic reaching the worker nodes on that port.