---
apiVersion: v1
kind: Pod
metadata:
  annotations:
    k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.132.0.39/23"],"mac_address":"0a:58:0a:84:00:27","gateway_ips":["10.132.0.1"],"routes":[{"dest":"10.132.0.0/14","nextHop":"10.132.0.1"},{"dest":"172.31.0.0/16","nextHop":"10.132.0.1"},{"dest":"169.254.0.5/32","nextHop":"10.132.0.1"},{"dest":"100.64.0.0/16","nextHop":"10.132.0.1"}],"ip_address":"10.132.0.39/23","gateway_ip":"10.132.0.1","role":"primary"}}'
    k8s.v1.cni.cncf.io/network-status: |-
      [{
          "name": "ovn-kubernetes",
          "interface": "eth0",
          "ips": [
              "10.132.0.39"
          ],
          "mac": "0a:58:0a:84:00:27",
          "default": true,
          "dns": {}
      }]
    openshift.io/scc: restricted-v2
    seccomp.security.alpha.kubernetes.io/pod: runtime/default
    security.openshift.io/validated-scc-subject-type: user
  creationTimestamp: "2026-04-24T17:13:34Z"
  generateName: stop-feature-test-kserve-router-scheduler-6c85574b4f-
  generation: 1
  labels:
    app.kubernetes.io/component: llminferenceservice-router-scheduler
    app.kubernetes.io/name: stop-feature-test
    app.kubernetes.io/part-of: llminferenceservice
    pod-template-hash: 6c85574b4f
  managedFields:
  - apiVersion: v1
    fieldsType: FieldsV1
    fieldsV1:
      f:metadata:
        f:annotations:
          f:k8s.ovn.org/pod-networks: {}
    manager: ip-10-0-137-69
    operation: Update
    subresource: status
    time: "2026-04-24T17:13:34Z"
  - apiVersion: v1
    fieldsType: FieldsV1
    fieldsV1:
      f:metadata:
        f:generateName: {}
        f:labels:
          .: {}
          f:app.kubernetes.io/component: {}
          f:app.kubernetes.io/name: {}
          f:app.kubernetes.io/part-of: {}
          f:pod-template-hash: {}
        f:ownerReferences:
          .: {}
          k:{"uid":"998579d2-3426-4020-a0c1-7956f2af0e04"}: {}
      f:spec:
        f:containers:
          k:{"name":"main"}:
            .: {}
            f:args: {}
            f:command: {}
            f:env:
              .: {}
              k:{"name":"SSL_CERT_DIR"}:
                .: {}
                f:name: {}
                f:value: {}
            f:image: {}
            f:imagePullPolicy: {}
            f:livenessProbe:
              .: {}
              f:failureThreshold: {}
              f:grpc:
                .: {}
                f:port: {}
                f:service: {}
              f:initialDelaySeconds: {}
              f:periodSeconds: {}
              f:successThreshold: {}
              f:timeoutSeconds: {}
            f:name: {}
            f:ports:
              .: {}
              k:{"containerPort":5557,"protocol":"TCP"}:
                .: {}
                f:containerPort: {}
                f:name: {}
                f:protocol: {}
              k:{"containerPort":9002,"protocol":"TCP"}:
                .: {}
                f:containerPort: {}
                f:name: {}
                f:protocol: {}
              k:{"containerPort":9003,"protocol":"TCP"}:
                .: {}
                f:containerPort: {}
                f:name: {}
                f:protocol: {}
              k:{"containerPort":9090,"protocol":"TCP"}:
                .: {}
                f:containerPort: {}
                f:name: {}
                f:protocol: {}
            f:readinessProbe:
              .: {}
              f:failureThreshold: {}
              f:grpc:
                .: {}
                f:port: {}
                f:service: {}
              f:initialDelaySeconds: {}
              f:periodSeconds: {}
              f:successThreshold: {}
              f:timeoutSeconds: {}
            f:resources:
              .: {}
              f:requests:
                .: {}
                f:cpu: {}
                f:memory: {}
            f:securityContext:
              .: {}
              f:allowPrivilegeEscalation: {}
              f:capabilities:
                .: {}
                f:drop: {}
              f:readOnlyRootFilesystem: {}
              f:runAsNonRoot: {}
              f:seccompProfile:
                .: {}
                f:type: {}
            f:terminationMessagePath: {}
            f:terminationMessagePolicy: {}
            f:volumeMounts:
              .: {}
              k:{"mountPath":"/tmp/tokenizer"}:
                .: {}
                f:mountPath: {}
                f:name: {}
              k:{"mountPath":"/var/run/kserve/tls"}:
                .: {}
                f:mountPath: {}
                f:name: {}
                f:readOnly: {}
          k:{"name":"tokenizer"}:
            .: {}
            f:env:
              .: {}
              k:{"name":"TOKENIZERS_DIR"}:
                .: {}
                f:name: {}
                f:value: {}
            f:image: {}
            f:imagePullPolicy: {}
            f:livenessProbe:
              .: {}
              f:failureThreshold: {}
              f:httpGet:
                .: {}
                f:path: {}
                f:port: {}
                f:scheme: {}
              f:periodSeconds: {}
              f:successThreshold: {}
              f:timeoutSeconds: {}
            f:name: {}
            f:ports:
              .: {}
              k:{"containerPort":8082,"protocol":"TCP"}:
                .: {}
                f:containerPort: {}
                f:name: {}
                f:protocol: {}
            f:readinessProbe:
              .: {}
              f:failureThreshold: {}
              f:httpGet:
                .: {}
                f:path: {}
                f:port: {}
                f:scheme: {}
              f:periodSeconds: {}
              f:successThreshold: {}
              f:timeoutSeconds: {}
            f:resources:
              .: {}
              f:requests:
                .: {}
                f:cpu: {}
                f:memory: {}
            f:securityContext:
              .: {}
              f:allowPrivilegeEscalation: {}
              f:capabilities:
                .: {}
                f:drop: {}
              f:readOnlyRootFilesystem: {}
              f:runAsNonRoot: {}
              f:seccompProfile:
                .: {}
                f:type: {}
            f:startupProbe:
              .: {}
              f:failureThreshold: {}
              f:httpGet:
                .: {}
                f:path: {}
                f:port: {}
                f:scheme: {}
              f:initialDelaySeconds: {}
              f:periodSeconds: {}
              f:successThreshold: {}
              f:timeoutSeconds: {}
            f:terminationMessagePath: {}
            f:terminationMessagePolicy: {}
            f:volumeMounts:
              .: {}
              k:{"mountPath":"/.cache"}:
                .: {}
                f:mountPath: {}
                f:name: {}
              k:{"mountPath":"/mnt/models/base"}:
                .: {}
                f:mountPath: {}
                f:name: {}
                f:readOnly: {}
              k:{"mountPath":"/tmp"}:
                .: {}
                f:mountPath: {}
                f:name: {}
              k:{"mountPath":"/tmp/tokenizer"}:
                .: {}
                f:mountPath: {}
                f:name: {}
            f:workingDir: {}
        f:dnsPolicy: {}
        f:enableServiceLinks: {}
        f:initContainers:
          .: {}
          k:{"name":"storage-initializer"}:
            .: {}
            f:args: {}
            f:env:
              .: {}
              k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}:
                .: {}
                f:name: {}
                f:value: {}
              k:{"name":"HF_XET_HIGH_PERFORMANCE"}:
                .: {}
                f:name: {}
                f:value: {}
              k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}:
                .: {}
                f:name: {}
                f:value: {}
              k:{"name":"STORAGE_ALLOW_PATTERNS"}:
                .: {}
                f:name: {}
                f:value: {}
            f:image: {}
            f:imagePullPolicy: {}
            f:name: {}
            f:resources:
              .: {}
              f:limits:
                .: {}
                f:cpu: {}
                f:memory: {}
              f:requests:
                .: {}
                f:cpu: {}
                f:memory: {}
            f:terminationMessagePath: {}
            f:terminationMessagePolicy: {}
            f:volumeMounts:
              .: {}
              k:{"mountPath":"/mnt/models"}:
                .: {}
                f:mountPath: {}
                f:name: {}
        f:restartPolicy: {}
        f:schedulerName: {}
        f:securityContext: {}
        f:serviceAccount: {}
        f:serviceAccountName: {}
        f:terminationGracePeriodSeconds: {}
        f:volumes:
          .: {}
          k:{"name":"kserve-provision-location"}:
            .: {}
            f:emptyDir: {}
            f:name: {}
          k:{"name":"tls-certs"}:
            .: {}
            f:name: {}
            f:secret:
              .: {}
              f:defaultMode: {}
              f:secretName: {}
          k:{"name":"tokenizer-cache"}:
            .: {}
            f:emptyDir: {}
            f:name: {}
          k:{"name":"tokenizer-tmp"}:
            .: {}
            f:emptyDir: {}
            f:name: {}
          k:{"name":"tokenizer-uds"}:
            .: {}
            f:emptyDir: {}
            f:name: {}
    manager: kube-controller-manager
    operation: Update
    time: "2026-04-24T17:13:34Z"
  - apiVersion: v1
    fieldsType: FieldsV1
    fieldsV1:
      f:metadata:
        f:annotations:
          f:k8s.v1.cni.cncf.io/network-status: {}
    manager: multus-daemon
    operation: Update
    subresource: status
    time: "2026-04-24T17:13:35Z"
  - apiVersion: v1
    fieldsType: FieldsV1
    fieldsV1:
      f:status:
        f:conditions:
          k:{"type":"ContainersReady"}:
            .: {}
            f:lastProbeTime: {}
            f:lastTransitionTime: {}
            f:status: {}
            f:type: {}
          k:{"type":"Initialized"}:
            .: {}
            f:lastProbeTime: {}
            f:lastTransitionTime: {}
            f:status: {}
            f:type: {}
          k:{"type":"PodReadyToStartContainers"}:
            .: {}
            f:lastProbeTime: {}
            f:lastTransitionTime: {}
            f:status: {}
            f:type: {}
          k:{"type":"Ready"}:
            .: {}
            f:lastProbeTime: {}
            f:lastTransitionTime: {}
            f:status: {}
            f:type: {}
        f:containerStatuses: {}
        f:hostIP: {}
        f:hostIPs: {}
        f:initContainerStatuses: {}
        f:phase: {}
        f:podIP: {}
        f:podIPs:
          .: {}
          k:{"ip":"10.132.0.39"}:
            .: {}
            f:ip: {}
        f:startTime: {}
    manager: kubelet
    operation: Update
    subresource: status
    time: "2026-04-24T17:14:06Z"
  name: stop-feature-test-kserve-router-scheduler-6c85574b4f-hkzx8
  namespace: kserve-ci-e2e-test
  ownerReferences:
  - apiVersion: apps/v1
    blockOwnerDeletion: true
    controller: true
    kind: ReplicaSet
    name: stop-feature-test-kserve-router-scheduler-6c85574b4f
    uid: 998579d2-3426-4020-a0c1-7956f2af0e04
  resourceVersion: "39686"
  uid: a31a70b8-65d8-418e-aa56-02717c6e7e73
spec:
  containers:
  - args:
    - --config-text
    - |
      apiVersion: inference.networking.x-k8s.io/v1alpha1
      kind: EndpointPickerConfig
      plugins:
      - type: single-profile-handler
      - type: queue-scorer
      - type: prefix-cache-scorer
      - type: max-score-picker
      schedulingProfiles:
      - name: default
        plugins:
        - pluginRef: queue-scorer
          weight: 2
        - pluginRef: prefix-cache-scorer
          weight: 3
        - pluginRef: max-score-picker
    command:
    - /app/epp
    - --pool-name
    - stop-feature-test-inference-pool
    - --pool-namespace
    - kserve-ci-e2e-test
    - --zap-encoder
    - json
    - --grpc-port
    - "9002"
    - --grpc-health-port
    - "9003"
    - --enable-cert-reload=true
    - --secure-serving=true
    - --model-server-metrics-scheme=https
    - --cert-path=/var/run/kserve/tls
    env:
    - name: SSL_CERT_DIR
      value: /var/run/kserve/tls:/var/run/secrets/kubernetes.io/serviceaccount:/etc/pki/tls/certs
    image: ghcr.io/llm-d/llm-d-inference-scheduler:v0.7.1
    imagePullPolicy: IfNotPresent
    livenessProbe:
      failureThreshold: 3
      grpc:
        port: 9003
        service: liveness
      initialDelaySeconds: 5
      periodSeconds: 10
      successThreshold: 1
      timeoutSeconds: 1
    name: main
    ports:
    - containerPort: 9002
      name: grpc
      protocol: TCP
    - containerPort: 9003
      name: grpc-health
      protocol: TCP
    - containerPort: 9090
      name: metrics
      protocol: TCP
    - containerPort: 5557
      name: zmq
      protocol: TCP
    readinessProbe:
      failureThreshold: 3
      grpc:
        port: 9003
        service: readiness
      initialDelaySeconds: 30
      periodSeconds: 10
      successThreshold: 1
      timeoutSeconds: 1
    resources:
      requests:
        cpu: 256m
        memory: 500Mi
    securityContext:
      allowPrivilegeEscalation: false
      capabilities:
        drop:
        - ALL
      readOnlyRootFilesystem: true
      runAsNonRoot: true
      runAsUser: 1000690000
      seccompProfile:
        type: RuntimeDefault
    terminationMessagePath: /dev/termination-log
    terminationMessagePolicy: FallbackToLogsOnError
    volumeMounts:
    - mountPath: /var/run/kserve/tls
      name: tls-certs
      readOnly: true
    - mountPath: /tmp/tokenizer
      name: tokenizer-uds
    - mountPath: /var/run/secrets/kubernetes.io/serviceaccount
      name: kube-api-access-wqvd7
      readOnly: true
  - env:
    - name: TOKENIZERS_DIR
      value: /mnt/models
    image: ghcr.io/llm-d/llm-d-uds-tokenizer:v0.7.1
    imagePullPolicy: IfNotPresent
    livenessProbe:
      failureThreshold: 3
      httpGet:
        path: /healthz
        port: 8082
        scheme: HTTP
      periodSeconds: 15
      successThreshold: 1
      timeoutSeconds: 5
    name: tokenizer
    ports:
    - containerPort: 8082
      name: health
      protocol: TCP
    readinessProbe:
      failureThreshold: 3
      httpGet:
        path: /healthz
        port: 8082
        scheme: HTTP
      periodSeconds: 10
      successThreshold: 1
      timeoutSeconds: 5
    resources:
      requests:
        cpu: 256m
        memory: 500Mi
    securityContext:
      allowPrivilegeEscalation: false
      capabilities:
        drop:
        - ALL
      readOnlyRootFilesystem: true
      runAsNonRoot: true
      runAsUser: 1000690000
      seccompProfile:
        type: RuntimeDefault
    startupProbe:
      failureThreshold: 60
      httpGet:
        path: /healthz
        port: 8082
        scheme: HTTP
      initialDelaySeconds: 5
      periodSeconds: 10
      successThreshold: 1
      timeoutSeconds: 5
    terminationMessagePath: /dev/termination-log
    terminationMessagePolicy: FallbackToLogsOnError
    volumeMounts:
    - mountPath: /tmp
      name: tokenizer-tmp
    - mountPath: /.cache
      name: tokenizer-cache
    - mountPath: /tmp/tokenizer
      name: tokenizer-uds
    - mountPath: /mnt/models/base
      name: kserve-provision-location
      readOnly: true
    - mountPath: /var/run/secrets/kubernetes.io/serviceaccount
      name: kube-api-access-wqvd7
      readOnly: true
    workingDir: /mnt/models
  dnsPolicy: ClusterFirst
  enableServiceLinks: true
  imagePullSecrets:
  - name: default-dockercfg-q89jh
  - name: stop-feature-test-epp-sa-dockercfg-hqqcv
  initContainers:
  - args:
    - hf://facebook/opt-125m
    - /mnt/models
    env:
    - name: HF_HUB_ENABLE_HF_TRANSFER
      value: "1"
    - name: HF_XET_HIGH_PERFORMANCE
      value: "1"
    - name: HF_XET_NUM_CONCURRENT_RANGE_GETS
      value: "8"
    - name: STORAGE_ALLOW_PATTERNS
      value: '["tokenizer.json", "tokenizer_config.json", "special_tokens_map.json",
        "vocab.json", "merges.txt", "config.json", "generation_config.json"]'
    image: quay.io/opendatahub/kserve-storage-initializer@sha256:e180c5801eecc9f4bbb52efa5a04a7e91697942e5c91b352ed2867858a22963a
    imagePullPolicy: IfNotPresent
    name: storage-initializer
    resources:
      limits:
        cpu: "1"
        memory: 24Gi
      requests:
        cpu: 100m
        memory: 100Mi
    securityContext:
      allowPrivilegeEscalation: false
      capabilities:
        drop:
        - ALL
      runAsNonRoot: true
      runAsUser: 1000690000
    terminationMessagePath: /dev/termination-log
    terminationMessagePolicy: FallbackToLogsOnError
    volumeMounts:
    - mountPath: /mnt/models
      name: kserve-provision-location
    - mountPath: /var/run/secrets/kubernetes.io/serviceaccount
      name: kube-api-access-wqvd7
      readOnly: true
  nodeName: ip-10-0-137-69.ec2.internal
  preemptionPolicy: PreemptLowerPriority
  priority: 0
  restartPolicy: Always
  schedulerName: default-scheduler
  securityContext:
    fsGroup: 1000690000
    seLinuxOptions:
      level: s0:c26,c20
    seccompProfile:
      type: RuntimeDefault
  serviceAccount: stop-feature-test-epp-sa
  serviceAccountName: stop-feature-test-epp-sa
  terminationGracePeriodSeconds: 30
  tolerations:
  - effect: NoExecute
    key: node.kubernetes.io/not-ready
    operator: Exists
    tolerationSeconds: 300
  - effect: NoExecute
    key: node.kubernetes.io/unreachable
    operator: Exists
    tolerationSeconds: 300
  - effect: NoSchedule
    key: node.kubernetes.io/memory-pressure
    operator: Exists
  volumes:
  - name: tls-certs
    secret:
      defaultMode: 420
      secretName: stop-feature-test-kserve-self-signed-certs
  - emptyDir: {}
    name: tokenizer-uds
  - emptyDir: {}
    name: tokenizer-tmp
  - emptyDir: {}
    name: tokenizer-cache
  - emptyDir: {}
    name: kserve-provision-location
  - name: kube-api-access-wqvd7
    projected:
      defaultMode: 420
      sources:
      - serviceAccountToken:
          expirationSeconds: 3607
          path: token
      - configMap:
          items:
          - key: ca.crt
            path: ca.crt
          name: kube-root-ca.crt
      - downwardAPI:
          items:
          - fieldRef:
              apiVersion: v1
              fieldPath: metadata.namespace
            path: namespace
      - configMap:
          items:
          - key: service-ca.crt
            path: service-ca.crt
          name: openshift-service-ca.crt
status:
  conditions:
  - lastProbeTime: null
    lastTransitionTime: "2026-04-24T17:13:35Z"
    status: "True"
    type: PodReadyToStartContainers
  - lastProbeTime: null
    lastTransitionTime: "2026-04-24T17:13:36Z"
    status: "True"
    type: Initialized
  - lastProbeTime: null
    lastTransitionTime: "2026-04-24T17:14:06Z"
    status: "True"
    type: Ready
  - lastProbeTime: null
    lastTransitionTime: "2026-04-24T17:14:06Z"
    status: "True"
    type: ContainersReady
  - lastProbeTime: null
    lastTransitionTime: "2026-04-24T17:13:34Z"
    status: "True"
    type: PodScheduled
  containerStatuses:
  - allocatedResources:
      cpu: 256m
      memory: 500Mi
    containerID: cri-o://78788a144f0e419cf41c1b614f9323139e0b498578f82daed7c94cfa64e3dd0d
    image: ghcr.io/llm-d/llm-d-inference-scheduler:v0.7.1
    imageID: ghcr.io/llm-d/llm-d-inference-scheduler@sha256:88de279c6eb6758a4c600de9730e49e46b04c392846afedd03d82447379c9e7a
    lastState: {}
    name: main
    ready: true
    resources:
      requests:
        cpu: 256m
        memory: 500Mi
    restartCount: 0
    started: true
    state:
      running:
        startedAt: "2026-04-24T17:13:36Z"
    user:
      linux:
        gid: 0
        supplementalGroups:
        - 0
        - 1000690000
        uid: 1000690000
    volumeMounts:
    - mountPath: /var/run/kserve/tls
      name: tls-certs
      readOnly: true
      recursiveReadOnly: Disabled
    - mountPath: /tmp/tokenizer
      name: tokenizer-uds
    - mountPath: /var/run/secrets/kubernetes.io/serviceaccount
      name: kube-api-access-wqvd7
      readOnly: true
      recursiveReadOnly: Disabled
  - allocatedResources:
      cpu: 256m
      memory: 500Mi
    containerID: cri-o://376d9725f52d70ae6d122c0e798091b6bb876f1d9da95fa7e22c8c912417b6f5
    image: ghcr.io/llm-d/llm-d-uds-tokenizer:v0.7.1
    imageID: ghcr.io/llm-d/llm-d-uds-tokenizer@sha256:aed091a51f3d64458f1fdb451d21f745186bb4517a7ba0c49913a0c617366a3e
    lastState: {}
    name: tokenizer
    ready: true
    resources:
      requests:
        cpu: 256m
        memory: 500Mi
    restartCount: 0
    started: true
    state:
      running:
        startedAt: "2026-04-24T17:13:36Z"
    user:
      linux:
        gid: 0
        supplementalGroups:
        - 0
        - 1000690000
        uid: 1000690000
    volumeMounts:
    - mountPath: /tmp
      name: tokenizer-tmp
    - mountPath: /.cache
      name: tokenizer-cache
    - mountPath: /tmp/tokenizer
      name: tokenizer-uds
    - mountPath: /mnt/models/base
      name: kserve-provision-location
      readOnly: true
      recursiveReadOnly: Disabled
    - mountPath: /var/run/secrets/kubernetes.io/serviceaccount
      name: kube-api-access-wqvd7
      readOnly: true
      recursiveReadOnly: Disabled
  hostIP: 10.0.137.69
  hostIPs:
  - ip: 10.0.137.69
  initContainerStatuses:
  - allocatedResources:
      cpu: 100m
      memory: 100Mi
    containerID: cri-o://92af2fde0f62e107fc682a12f873b950bfee813bb27dfbc16a6a4df4022e9ab4
    image: quay.io/opendatahub/kserve-storage-initializer@sha256:e180c5801eecc9f4bbb52efa5a04a7e91697942e5c91b352ed2867858a22963a
    imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:beb2d86857b6be056be5b9b8be9f2d0cae6bec5194465a287ac65bbb72e60fd0
    lastState: {}
    name: storage-initializer
    ready: true
    resources:
      limits:
        cpu: "1"
        memory: 24Gi
      requests:
        cpu: 100m
        memory: 100Mi
    restartCount: 0
    started: false
    state:
      terminated:
        containerID: cri-o://92af2fde0f62e107fc682a12f873b950bfee813bb27dfbc16a6a4df4022e9ab4
        exitCode: 0
        finishedAt: "2026-04-24T17:13:35Z"
        reason: Completed
        startedAt: "2026-04-24T17:13:35Z"
    user:
      linux:
        gid: 0
        supplementalGroups:
        - 0
        - 1000690000
        uid: 1000690000
    volumeMounts:
    - mountPath: /mnt/models
      name: kserve-provision-location
    - mountPath: /var/run/secrets/kubernetes.io/serviceaccount
      name: kube-api-access-wqvd7
      readOnly: true
      recursiveReadOnly: Disabled
  phase: Running
  podIP: 10.132.0.39
  podIPs:
  - ip: 10.132.0.39
  qosClass: Burstable
  startTime: "2026-04-24T17:13:34Z"