# kube-loxilb path (guide section 4.2): the InferencePool exposed through a Gateway listener.
#
# kube-loxilb (--gatewayAPI --inferenceExtension) watches these three objects plus the pool in
# 30-inferencepool.yaml and creates the LoadBalancer Service "pd-disaggregation-inference", which it
# then programs on loxilb-igw as the same rule deploy/epp/rule.sh would create. Where each value comes from:
#
#   rule / Service field                        source
#   ------------------------------------------- ---------------------------------------------
#   namespace                                   the InferencePool's namespace
#   externalIP                                  Gateway spec.addresses, or an address from --cidrPools
#   port                                        Gateway listener port
#   targetPort                                  InferencePool spec.targetPorts[0]
#   endpoints (pod selector)                    InferencePool spec.selector.matchLabels
#   loxilb.io/epp-endpoint <ClusterIP>:9002     endpointPickerRef (name, port) -> Service ClusterIP:port
#   loxilb.io/epp-failure-mode                  endpointPickerRef.failureMode
#   loxilb.io/lbmode fullproxy                  kube-loxilb's default for a pool
#   loxilb.io/usepodnetwork "yes"               always set by kube-loxilb
#   loadBalancerClass loxilb.io/loxilb          kube-loxilb --loxilbLoadBalancerClass (default)
#
# Prerequisites: kube-loxilb running with --gatewayAPI --inferenceExtension; Gateway API CRDs installed.
apiVersion: gateway.networking.k8s.io/v1
kind: GatewayClass
metadata:
  name: loxilb-gc
spec:
  controllerName: loxilb.io/loxilb
---
apiVersion: gateway.networking.k8s.io/v1
kind: Gateway
metadata:
  name: inference-gw
  namespace: llm-d-pd-disaggregation
spec:
  gatewayClassName: loxilb-gc
  # No addresses: kube-loxilb assigns one from --cidrPools (the gateway node's IP in this setup).
  # To pin one:
  # addresses:
  #   - type: IPAddress
  #     value: 192.0.2.10
  listeners:
    - name: http
      protocol: HTTP
      port: 8080
      allowedRoutes:
        namespaces:
          from: Same
---
apiVersion: gateway.networking.k8s.io/v1
kind: HTTPRoute
metadata:
  name: llm-route
  namespace: llm-d-pd-disaggregation
spec:
  parentRefs:
    - name: inference-gw
      sectionName: http
  rules:
    - backendRefs:
        - group: inference.networking.k8s.io
          kind: InferencePool
          name: pd-disaggregation