--- apiVersion: v1 items: - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.134.0.43/23"],"mac_address":"0a:58:0a:86:00:2b","gateway_ips":["10.134.0.1"],"routes":[{"dest":"10.132.0.0/14","nextHop":"10.134.0.1"},{"dest":"172.31.0.0/16","nextHop":"10.134.0.1"},{"dest":"169.254.0.5/32","nextHop":"10.134.0.1"},{"dest":"100.64.0.0/16","nextHop":"10.134.0.1"}],"ip_address":"10.134.0.43/23","gateway_ip":"10.134.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.134.0.43" ], "mac": "0a:58:0a:86:00:2b", "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: "2026-07-09T18:31:11Z" generateName: e2e-pvc-model-download- generation: 1 labels: batch.kubernetes.io/controller-uid: 4c22ced8-0977-4835-a250-fea68542abd1 batch.kubernetes.io/job-name: e2e-pvc-model-download controller-uid: 4c22ced8-0977-4835-a250-fea68542abd1 job-name: e2e-pvc-model-download managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:generateName: {} f:labels: .: {} f:batch.kubernetes.io/controller-uid: {} f:batch.kubernetes.io/job-name: {} f:controller-uid: {} f:job-name: {} f:ownerReferences: .: {} k:{"uid":"4c22ced8-0977-4835-a250-fea68542abd1"}: {} f:spec: f:containers: k:{"name":"storage-initializer"}: .: {} f:args: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:dnsPolicy: {} f:enableServiceLinks: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"model-storage"}: .: {} f:name: {} f:persistentVolumeClaim: .: {} f:claimName: {} manager: kube-controller-manager operation: Update time: "2026-07-09T18:31:11Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-129-87 operation: Update subresource: status time: "2026-07-09T18:31:14Z" - 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: "2026-07-09T18:31:17Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:status: f:conditions: k:{"type":"ContainersReady"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:reason: {} f:status: {} f:type: {} k:{"type":"Initialized"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:reason: {} f:status: {} f:type: {} k:{"type":"PodReadyToStartContainers"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodScheduled"}: f:observedGeneration: {} k:{"type":"Ready"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:reason: {} f:status: {} f:type: {} f:containerStatuses: {} f:hostIP: {} f:hostIPs: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.134.0.43"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-09T18:31:25Z" name: e2e-pvc-model-download-hx52p namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: batch/v1 blockOwnerDeletion: true controller: true kind: Job name: e2e-pvc-model-download uid: 4c22ced8-0977-4835-a250-fea68542abd1 resourceVersion: "35790" uid: d1a50ed2-da02-4097-8351-d11e25d71cb9 spec: containers: - args: - hf://facebook/opt-125m - /mnt/models image: quay.io/opendatahub/kserve-storage-initializer@sha256:2cc17867b5b866a90feccd95d1b7d2c2c966c59c496775c7d4238555c4abef16 imagePullPolicy: IfNotPresent name: storage-initializer resources: {} securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: File volumeMounts: - mountPath: /mnt/models name: model-storage - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-qk5qc readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-s8z92 nodeName: ip-10-0-129-87.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Never schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: default serviceAccountName: default 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 volumes: - name: model-storage persistentVolumeClaim: claimName: e2e-pvc-model-storage - name: kube-api-access-qk5qc 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: "2026-07-09T18:31:25Z" observedGeneration: 1 status: "False" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-09T18:31:14Z" observedGeneration: 1 reason: PodCompleted status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-09T18:31:24Z" observedGeneration: 1 reason: PodCompleted status: "False" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-09T18:31:24Z" observedGeneration: 1 reason: PodCompleted status: "False" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-09T18:31:14Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - containerID: cri-o://982fd4bb11846f8827eaec352c1ffd71288a6c22b45a90660152218571cd03f6 image: quay.io/opendatahub/kserve-storage-initializer@sha256:2cc17867b5b866a90feccd95d1b7d2c2c966c59c496775c7d4238555c4abef16 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:2cc17867b5b866a90feccd95d1b7d2c2c966c59c496775c7d4238555c4abef16 lastState: {} name: storage-initializer ready: false resources: {} restartCount: 0 started: false state: terminated: containerID: cri-o://982fd4bb11846f8827eaec352c1ffd71288a6c22b45a90660152218571cd03f6 exitCode: 0 finishedAt: "2026-07-09T18:31:24Z" reason: Completed startedAt: "2026-07-09T18:31:17Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: model-storage - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-qk5qc readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.129.87 hostIPs: - ip: 10.0.129.87 observedGeneration: 1 phase: Succeeded podIP: 10.134.0.43 podIPs: - ip: 10.134.0.43 qosClass: BestEffort startTime: "2026-07-09T18:31:14Z" - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.132.0.42/23"],"mac_address":"0a:58:0a:84:00:2a","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":"10.132.0.42/23","gateway_ip":"10.132.0.1","role":"primary"}}' openshift.io/scc: restricted-v2 seccomp.security.alpha.kubernetes.io/pod: runtime/default security.openshift.io/validated-scc-subject-type: user creationTimestamp: "2026-07-09T18:31:36Z" generateName: llmisvc-model-pvc-router-manage-59b9d263-kserve-585587bc9d- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload app.kubernetes.io/name: llmisvc-model-pvc-router-manage-59b9d263 app.kubernetes.io/part-of: llminferenceservice kserve.io/component: workload llm-d.ai/role: both pod-template-hash: 585587bc9d managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-137-79 operation: Update subresource: status time: "2026-07-09T18:31:36Z" - 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:kserve.io/component: {} f:llm-d.ai/role: {} f:pod-template-hash: {} f:ownerReferences: .: {} k:{"uid":"9233eca4-c8c6-4599-b0f0-d8eb45685154"}: {} f:spec: f:containers: k:{"name":"main"}: .: {} f:command: {} f:env: .: {} k:{"name":"HF_HUB_CACHE"}: .: {} f:name: {} f:value: {} k:{"name":"HOME"}: .: {} f:name: {} f:value: {} k:{"name":"TORCHINDUCTOR_CACHE_DIR"}: .: {} f:name: {} f:value: {} k:{"name":"USER"}: .: {} f:name: {} f:value: {} k:{"name":"VLLM_CPU_KVCACHE_SPACE"}: .: {} f:name: {} f:value: {} k:{"name":"VLLM_ENABLE_V1_MULTIPROCESSING"}: .: {} f:name: {} f:value: {} k:{"name":"VLLM_LOGGING_LEVEL"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:lifecycle: .: {} f:preStop: .: {} f:exec: .: {} f:command: {} f:livenessProbe: .: {} f:failureThreshold: {} f:httpGet: .: {} f:path: {} f:port: {} f:scheme: {} f:periodSeconds: {} f:successThreshold: {} f:timeoutSeconds: {} f:name: {} f:ports: .: {} k:{"containerPort":8000,"protocol":"TCP"}: .: {} f:containerPort: {} f:protocol: {} f:readinessProbe: .: {} f:failureThreshold: {} f:httpGet: .: {} f:path: {} f:port: {} f:scheme: {} f:periodSeconds: {} f:successThreshold: {} f:timeoutSeconds: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} 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:periodSeconds: {} f:successThreshold: {} f:timeoutSeconds: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/dev/shm"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/home"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:readOnly: {} k:{"mountPath":"/models"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/tmp"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/kserve/tls"}: .: {} f:mountPath: {} f:name: {} f:readOnly: {} f:dnsPolicy: {} f:enableServiceLinks: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"dshm"}: .: {} f:emptyDir: .: {} f:medium: {} f:sizeLimit: {} f:name: {} k:{"name":"home"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"kserve-pvc-source"}: .: {} f:name: {} f:persistentVolumeClaim: .: {} f:claimName: {} k:{"name":"model-cache"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tls-certs"}: .: {} f:name: {} f:secret: .: {} f:defaultMode: {} f:secretName: {} k:{"name":"tmp-dir"}: .: {} f:emptyDir: {} f:name: {} manager: kube-controller-manager operation: Update time: "2026-07-09T18:31:36Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:status: f:conditions: k:{"type":"ContainersReady"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:message: {} f:observedGeneration: {} f:reason: {} f:status: {} f:type: {} k:{"type":"Initialized"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodReadyToStartContainers"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodScheduled"}: f:observedGeneration: {} k:{"type":"Ready"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:message: {} f:observedGeneration: {} f:reason: {} f:status: {} f:type: {} f:containerStatuses: {} f:hostIP: {} f:hostIPs: {} f:observedGeneration: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-09T18:31:36Z" name: llmisvc-model-pvc-router-manage-59b9d263-kserve-585587bc9d2pjx4 namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: llmisvc-model-pvc-router-manage-59b9d263-kserve-585587bc9d uid: 9233eca4-c8c6-4599-b0f0-d8eb45685154 resourceVersion: "35923" uid: a536c660-613b-4de7-9415-ccdf91ae856f spec: containers: - command: - /bin/bash - -c - |- if [ -f /etc/profile.d/ibm-aiu-setup.sh ]; then source /etc/profile.d/ibm-aiu-setup.sh fi if [ "$KSERVE_INFER_ROCE" = "true" ]; then echo "Trying to infer RoCE configs ... " grep -H . /sys/class/infiniband/*/ports/*/gids/* 2>/dev/null grep -H . /sys/class/infiniband/*/ports/*/gid_attrs/types/* 2>/dev/null cat /proc/driver/nvidia/params KSERVE_INFER_IB_GID_INDEX_GREP=${KSERVE_INFER_IB_GID_INDEX_GREP:-"RoCE v2"} echo "[Infer RoCE] Discovering active HCAs ..." active_hcas=() # Loop through all mlx5 devices found in sysfs for hca_dir in /sys/class/infiniband/mlx5_*; do # Ensure it's a directory before proceeding if [ -d "$hca_dir" ]; then hca_name=$(basename "$hca_dir") port_state_file="$hca_dir/ports/1/state" # Assume port 1 type_file="$hca_dir/ports/1/gid_attrs/types/*" echo "[Infer RoCE] Check if the port state file ${port_state_file} exists and contains 'ACTIVE'" if [ -f "$port_state_file" ] && grep -q "ACTIVE" "$port_state_file" && grep -q "${KSERVE_INFER_IB_GID_INDEX_GREP}" ${type_file} 2>/dev/null; then echo "[Infer RoCE] Found active HCA: $hca_name" active_hcas+=("$hca_name") else echo "[Infer RoCE] Skipping inactive or down HCA: $hca_name" fi fi done # Check if we found any active HCAs if [ ${#active_hcas[@]} -gt 0 ]; then # Join the array elements with a comma hca_port_pairs=() for hca in "${active_hcas[@]}"; do hca_port_pairs+=("${hca}:1") done active_hca_list=$(IFS=,; echo "${active_hcas[*]}") hca_port_pairs_list=$(IFS=,; echo "${hca_port_pairs[*]}") echo "[Infer RoCE] Setting active HCAs: ${active_hca_list}" export NCCL_IB_HCA=${NCCL_IB_HCA:-${active_hca_list}} export NVSHMEM_HCA_LIST=${NVSHMEM_HCA_LIST:-${hca_port_pairs_list}} export UCX_NET_DEVICES=${UCX_NET_DEVICES:-${hca_port_pairs_list}} echo "[Infer RoCE] NCCL_IB_HCA=${NCCL_IB_HCA}" echo "[Infer RoCE] NVSHMEM_HCA_LIST=${NVSHMEM_HCA_LIST}" echo "[Infer RoCE] UCX_NET_DEVICES=${UCX_NET_DEVICES}" else echo "[Infer RoCE] WARNING: No active RoCE HCAs found. NCCL_IB_HCA will not be set." fi if [ ${#active_hcas[@]} -gt 0 ]; then echo "[Infer RoCE] Finding GID_INDEX for each active HCA (SR-IOV compatible)..." # For SR-IOV environments, find the most common IPv4 RoCE v2 GID index across all HCAs declare -A gid_index_count declare -A hca_gid_index for hca_name in "${active_hcas[@]}"; do echo "[Infer RoCE] Processing HCA: ${hca_name}" # Find all RoCE v2 IPv4 GIDs for this HCA and count by index for tpath in /sys/class/infiniband/${hca_name}/ports/1/gid_attrs/types/*; do if grep -q "${KSERVE_INFER_IB_GID_INDEX_GREP}" "$tpath" 2>/dev/null; then idx=$(basename "$tpath") gid_file="/sys/class/infiniband/${hca_name}/ports/1/gids/${idx}" # Check for IPv4 GID (contains ffff:) if [ -f "$gid_file" ] && grep -q "ffff:" "$gid_file"; then gid_value=$(cat "$gid_file" 2>/dev/null || echo "") echo "[Infer RoCE] Found IPv4 RoCE v2 GID for ${hca_name}: index=${idx}, gid=${gid_value}" hca_gid_index["${hca_name}"]="${idx}" gid_index_count["${idx}"]=$((${gid_index_count["${idx}"]} + 1)) break # Use first found IPv4 GID per HCA fi fi done done # Find the most common GID index (most likely to be consistent across nodes) best_gid_index="" max_count=0 for idx in "${!gid_index_count[@]}"; do count=${gid_index_count["${idx}"]} echo "[Infer RoCE] GID_INDEX ${idx} found on ${count} HCAs" if [ $count -gt $max_count ]; then max_count=$count best_gid_index="$idx" fi done # Use deterministic fallback if tied - prefer index 3 (SR-IOV standard) if [ ${#gid_index_count[@]} -gt 1 ]; then echo "[Infer RoCE] Multiple GID indices found, selecting most common: ${best_gid_index}" # If there's a tie, prefer index 3 as it's most common in SR-IOV setups if [ -n "${gid_index_count['3']}" ] && [ "${gid_index_count['3']}" -eq "$max_count" ]; then best_gid_index="3" echo "[Infer RoCE] Using deterministic fallback: GID_INDEX=3 (SR-IOV standard)" fi fi # Check if GID_INDEX is already set via environment variables if [ -n "${NCCL_IB_GID_INDEX}" ]; then echo "[Infer RoCE] Using pre-configured NCCL_IB_GID_INDEX=${NCCL_IB_GID_INDEX} from environment" export NVSHMEM_IB_GID_INDEX=${NVSHMEM_IB_GID_INDEX:-$NCCL_IB_GID_INDEX} export UCX_IB_GID_INDEX=${UCX_IB_GID_INDEX:-$NCCL_IB_GID_INDEX} echo "[Infer RoCE] Using pre-configured GID_INDEX=${NCCL_IB_GID_INDEX} for NCCL, NVSHMEM, and UCX" elif [ -n "$best_gid_index" ]; then echo "[Infer RoCE] Selected GID_INDEX: ${best_gid_index} (found on ${max_count} HCAs)" export NCCL_IB_GID_INDEX=${NCCL_IB_GID_INDEX:-$best_gid_index} export NVSHMEM_IB_GID_INDEX=${NVSHMEM_IB_GID_INDEX:-$best_gid_index} export UCX_IB_GID_INDEX=${UCX_IB_GID_INDEX:-$best_gid_index} echo "[Infer RoCE] Exported GID_INDEX=${best_gid_index} for NCCL, NVSHMEM, and UCX" else echo "[Infer RoCE] ERROR: No valid IPv4 ${KSERVE_INFER_IB_GID_INDEX_GREP} GID_INDEX found on any HCA." fi else echo "[Infer RoCE] No active HCAs found, skipping GID_INDEX inference." fi fi # --disable-access-log-for-endpoints landed in vLLM 0.16.0 (vllm-project/vllm#30011). # Older versions still need the blanket --disable-uvicorn-access-log. ACCESS_LOG_ARGS="--disable-uvicorn-access-log" VLLM_VERSION=$(vllm --version 2>/dev/null | tail -1 | awk '{print $NF}') echo "[access-log-detect] vllm version='${VLLM_VERSION}'" if [[ "$VLLM_VERSION" =~ ^[0-9]+\.[0-9]+ ]] && [ "$(printf '%s\n%s\n' "0.16.0" "${VLLM_VERSION}" | sort -V | head -1)" = "0.16.0" ]; then ACCESS_LOG_ARGS="--disable-access-log-for-endpoints /health,/metrics,/ping" fi echo "[access-log-detect] selected ACCESS_LOG_ARGS='${ACCESS_LOG_ARGS}'" # --shutdown-timeout landed in vLLM 0.18.0 (vllm-project/vllm#36666). SHUTDOWN_TIMEOUT_ARGS="" if [[ "$VLLM_VERSION" =~ ^[0-9]+\.[0-9]+ ]] && [ "$(printf '%s\n%s\n' "0.18.0" "${VLLM_VERSION}" | sort -V | head -1)" = "0.18.0" ]; then SHUTDOWN_TIMEOUT_ARGS="--shutdown-timeout 40" fi eval "exec vllm serve /mnt/models \ --served-model-name "facebook/opt-125m" "publishers/kserve-ci-e2e-test/models/facebook/opt-125m" \ --port 8000 \ ${ACCESS_LOG_ARGS} \ ${SHUTDOWN_TIMEOUT_ARGS} \ --enable-ssl-refresh \ --ssl-certfile /var/run/kserve/tls/tls.crt \ --ssl-keyfile /var/run/kserve/tls/tls.key \ ${VLLM_ADDITIONAL_ARGS} \ $@" - -- env: - name: HOME value: /home - name: VLLM_LOGGING_LEVEL value: DEBUG - name: VLLM_CPU_KVCACHE_SPACE value: "1" - name: VLLM_ENABLE_V1_MULTIPROCESSING value: "0" - name: USER value: nonroot - name: TORCHINDUCTOR_CACHE_DIR value: /tmp/torchinductor-cache - name: HF_HUB_CACHE value: /models image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imagePullPolicy: IfNotPresent lifecycle: preStop: exec: command: - /bin/sleep - "15" livenessProbe: failureThreshold: 10 httpGet: path: /health port: 8000 scheme: HTTPS periodSeconds: 10 successThreshold: 1 timeoutSeconds: 1 name: main ports: - containerPort: 8000 protocol: TCP readinessProbe: failureThreshold: 2 httpGet: path: /health port: 8000 scheme: HTTPS periodSeconds: 1 successThreshold: 1 timeoutSeconds: 1 resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL readOnlyRootFilesystem: true runAsNonRoot: true runAsUser: 1000690000 seccompProfile: type: RuntimeDefault startupProbe: failureThreshold: 60 httpGet: path: /health port: 8000 scheme: HTTPS periodSeconds: 10 successThreshold: 1 timeoutSeconds: 1 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /home name: home - mountPath: /tmp name: tmp-dir - mountPath: /dev/shm name: dshm - mountPath: /models name: model-cache - mountPath: /var/run/kserve/tls name: tls-certs readOnly: true - mountPath: /mnt/models name: kserve-pvc-source readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-fsw26 readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-s8z92 nodeName: ip-10-0-137-79.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: default serviceAccountName: default terminationGracePeriodSeconds: 60 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: - emptyDir: {} name: home - emptyDir: medium: Memory sizeLimit: 1Gi name: dshm - emptyDir: {} name: model-cache - emptyDir: {} name: tmp-dir - name: tls-certs secret: defaultMode: 420 secretName: llmisv8bf079eb6eda4debfb4ef5bb7817824c-kserve-self-signed-certs - name: kserve-pvc-source persistentVolumeClaim: claimName: e2e-pvc-model-storage - name: kube-api-access-fsw26 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: "2026-07-09T18:31:36Z" observedGeneration: 1 status: "False" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-09T18:31:36Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-09T18:31:36Z" message: 'containers with unready status: [main]' observedGeneration: 1 reason: ContainersNotReady status: "False" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-09T18:31:36Z" message: 'containers with unready status: [main]' observedGeneration: 1 reason: ContainersNotReady status: "False" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-09T18:31:36Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: "" lastState: {} name: main ready: false restartCount: 0 started: false state: waiting: reason: ContainerCreating volumeMounts: - mountPath: /home name: home - mountPath: /tmp name: tmp-dir - mountPath: /dev/shm name: dshm - mountPath: /models name: model-cache - mountPath: /var/run/kserve/tls name: tls-certs readOnly: true recursiveReadOnly: Disabled - mountPath: /mnt/models name: kserve-pvc-source readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-fsw26 readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.137.79 hostIPs: - ip: 10.0.137.79 observedGeneration: 1 phase: Pending qosClass: Burstable startTime: "2026-07-09T18:31:36Z" - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.132.0.51/23"],"mac_address":"0a:58:0a:84:00:33","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":"10.132.0.51/23","gateway_ip":"10.132.0.1","role":"primary"}}' openshift.io/scc: restricted-v2 seccomp.security.alpha.kubernetes.io/pod: runtime/default security.openshift.io/validated-scc-subject-type: user creationTimestamp: "2026-07-09T18:46:57Z" generateName: llmisvc-model-pvc-router-manage-e8706282-kserve-7db79954c- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload app.kubernetes.io/name: llmisvc-model-pvc-router-manage-e8706282 app.kubernetes.io/part-of: llminferenceservice kserve.io/component: workload llm-d.ai/role: decode pod-template-hash: 7db79954c managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-137-79 operation: Update subresource: status time: "2026-07-09T18:46:57Z" - 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:kserve.io/component: {} f:llm-d.ai/role: {} f:pod-template-hash: {} f:ownerReferences: .: {} k:{"uid":"e0453e37-f07d-4d38-86e3-0283a1dc1755"}: {} f:spec: f:containers: k:{"name":"main"}: .: {} f:command: {} f:env: .: {} k:{"name":"HF_HUB_CACHE"}: .: {} f:name: {} f:value: {} k:{"name":"HOME"}: .: {} f:name: {} f:value: {} k:{"name":"TORCHINDUCTOR_CACHE_DIR"}: .: {} f:name: {} f:value: {} k:{"name":"USER"}: .: {} f:name: {} f:value: {} k:{"name":"VLLM_CPU_KVCACHE_SPACE"}: .: {} f:name: {} f:value: {} k:{"name":"VLLM_ENABLE_V1_MULTIPROCESSING"}: .: {} f:name: {} f:value: {} k:{"name":"VLLM_LOGGING_LEVEL"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:lifecycle: .: {} f:preStop: .: {} f:exec: .: {} f:command: {} f:livenessProbe: .: {} f:failureThreshold: {} f:httpGet: .: {} f:path: {} f:port: {} f:scheme: {} f:initialDelaySeconds: {} f:periodSeconds: {} f:successThreshold: {} f:timeoutSeconds: {} f:name: {} f:ports: .: {} k:{"containerPort":8001,"protocol":"TCP"}: .: {} f:containerPort: {} f:protocol: {} f:readinessProbe: .: {} f:failureThreshold: {} f:httpGet: .: {} f:path: {} f:port: {} f:scheme: {} f:initialDelaySeconds: {} f:periodSeconds: {} f:successThreshold: {} f:timeoutSeconds: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} 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:periodSeconds: {} f:successThreshold: {} f:timeoutSeconds: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/dev/shm"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/home"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:readOnly: {} k:{"mountPath":"/models"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/tmp"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/kserve/tls"}: .: {} f:mountPath: {} f:name: {} f:readOnly: {} f:dnsPolicy: {} f:enableServiceLinks: {} f:initContainers: .: {} k:{"name":"llm-d-routing-sidecar"}: .: {} f:command: {} f:env: .: {} k:{"name":"INFERENCE_POOL_NAME"}: .: {} f:name: {} f:value: {} k:{"name":"INFERENCE_POOL_NAMESPACE"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"SSL_CERT_DIR"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:livenessProbe: .: {} f:failureThreshold: {} f:httpGet: .: {} f:path: {} f:port: {} f:scheme: {} f:initialDelaySeconds: {} f:periodSeconds: {} f:successThreshold: {} f:timeoutSeconds: {} f:name: {} f:ports: .: {} k:{"containerPort":8000,"protocol":"TCP"}: .: {} f:containerPort: {} f:protocol: {} f:readinessProbe: .: {} f:failureThreshold: {} f:httpGet: .: {} f:path: {} f:port: {} f:scheme: {} f:initialDelaySeconds: {} f:periodSeconds: {} f:successThreshold: {} f:timeoutSeconds: {} f:resources: {} f:restartPolicy: {} f:securityContext: .: {} f:allowPrivilegeEscalation: {} f:capabilities: .: {} f:drop: {} f:readOnlyRootFilesystem: {} f:runAsNonRoot: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/var/run/kserve/tls"}: .: {} f:mountPath: {} f:name: {} f:readOnly: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"dshm"}: .: {} f:emptyDir: .: {} f:medium: {} f:sizeLimit: {} f:name: {} k:{"name":"home"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"kserve-pvc-source"}: .: {} f:name: {} f:persistentVolumeClaim: .: {} f:claimName: {} k:{"name":"model-cache"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tls-certs"}: .: {} f:name: {} f:secret: .: {} f:defaultMode: {} f:secretName: {} k:{"name":"tmp-dir"}: .: {} f:emptyDir: {} f:name: {} manager: kube-controller-manager operation: Update time: "2026-07-09T18:46:57Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:status: f:conditions: k:{"type":"ContainersReady"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:message: {} f:observedGeneration: {} f:reason: {} f:status: {} f:type: {} k:{"type":"Initialized"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:message: {} f:observedGeneration: {} f:reason: {} f:status: {} f:type: {} k:{"type":"PodReadyToStartContainers"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodScheduled"}: f:observedGeneration: {} k:{"type":"Ready"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:message: {} f:observedGeneration: {} f:reason: {} f:status: {} f:type: {} f:containerStatuses: {} f:hostIP: {} f:hostIPs: {} f:initContainerStatuses: {} f:observedGeneration: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-09T18:46:57Z" name: llmisvc-model-pvc-router-manage-e8706282-kserve-7db79954c-nwr94 namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: llmisvc-model-pvc-router-manage-e8706282-kserve-7db79954c uid: e0453e37-f07d-4d38-86e3-0283a1dc1755 resourceVersion: "47430" uid: f790da47-896e-405c-ba3f-608321ec51e5 spec: containers: - command: - /bin/bash - -c - |- if [ -f /etc/profile.d/ibm-aiu-setup.sh ]; then source /etc/profile.d/ibm-aiu-setup.sh fi if [ "$KSERVE_INFER_ROCE" = "true" ]; then echo "Trying to infer RoCE configs ... " grep -H . /sys/class/infiniband/*/ports/*/gids/* 2>/dev/null grep -H . /sys/class/infiniband/*/ports/*/gid_attrs/types/* 2>/dev/null cat /proc/driver/nvidia/params KSERVE_INFER_IB_GID_INDEX_GREP=${KSERVE_INFER_IB_GID_INDEX_GREP:-"RoCE v2"} echo "[Infer RoCE] Discovering active HCAs ..." active_hcas=() # Loop through all mlx5 devices found in sysfs for hca_dir in /sys/class/infiniband/mlx5_*; do # Ensure it's a directory before proceeding if [ -d "$hca_dir" ]; then hca_name=$(basename "$hca_dir") port_state_file="$hca_dir/ports/1/state" # Assume port 1 type_file="$hca_dir/ports/1/gid_attrs/types/*" echo "[Infer RoCE] Check if the port state file ${port_state_file} exists and contains 'ACTIVE'" if [ -f "$port_state_file" ] && grep -q "ACTIVE" "$port_state_file" && grep -q "${KSERVE_INFER_IB_GID_INDEX_GREP}" ${type_file} 2>/dev/null; then echo "[Infer RoCE] Found active HCA: $hca_name" active_hcas+=("$hca_name") else echo "[Infer RoCE] Skipping inactive or down HCA: $hca_name" fi fi done # Check if we found any active HCAs if [ ${#active_hcas[@]} -gt 0 ]; then # Join the array elements with a comma hca_port_pairs=() for hca in "${active_hcas[@]}"; do hca_port_pairs+=("${hca}:1") done active_hca_list=$(IFS=,; echo "${active_hcas[*]}") hca_port_pairs_list=$(IFS=,; echo "${hca_port_pairs[*]}") echo "[Infer RoCE] Setting active HCAs: ${active_hca_list}" export NCCL_IB_HCA=${NCCL_IB_HCA:-${active_hca_list}} export NVSHMEM_HCA_LIST=${NVSHMEM_HCA_LIST:-${hca_port_pairs_list}} export UCX_NET_DEVICES=${UCX_NET_DEVICES:-${hca_port_pairs_list}} echo "[Infer RoCE] NCCL_IB_HCA=${NCCL_IB_HCA}" echo "[Infer RoCE] NVSHMEM_HCA_LIST=${NVSHMEM_HCA_LIST}" echo "[Infer RoCE] UCX_NET_DEVICES=${UCX_NET_DEVICES}" else echo "[Infer RoCE] WARNING: No active RoCE HCAs found. NCCL_IB_HCA will not be set." fi if [ ${#active_hcas[@]} -gt 0 ]; then echo "[Infer RoCE] Finding GID_INDEX for each active HCA (SR-IOV compatible)..." # For SR-IOV environments, find the most common IPv4 RoCE v2 GID index across all HCAs declare -A gid_index_count declare -A hca_gid_index for hca_name in "${active_hcas[@]}"; do echo "[Infer RoCE] Processing HCA: ${hca_name}" # Find all RoCE v2 IPv4 GIDs for this HCA and count by index for tpath in /sys/class/infiniband/${hca_name}/ports/1/gid_attrs/types/*; do if grep -q "${KSERVE_INFER_IB_GID_INDEX_GREP}" "$tpath" 2>/dev/null; then idx=$(basename "$tpath") gid_file="/sys/class/infiniband/${hca_name}/ports/1/gids/${idx}" # Check for IPv4 GID (contains ffff:) if [ -f "$gid_file" ] && grep -q "ffff:" "$gid_file"; then gid_value=$(cat "$gid_file" 2>/dev/null || echo "") echo "[Infer RoCE] Found IPv4 RoCE v2 GID for ${hca_name}: index=${idx}, gid=${gid_value}" hca_gid_index["${hca_name}"]="${idx}" gid_index_count["${idx}"]=$((${gid_index_count["${idx}"]} + 1)) break # Use first found IPv4 GID per HCA fi fi done done # Find the most common GID index (most likely to be consistent across nodes) best_gid_index="" max_count=0 for idx in "${!gid_index_count[@]}"; do count=${gid_index_count["${idx}"]} echo "[Infer RoCE] GID_INDEX ${idx} found on ${count} HCAs" if [ $count -gt $max_count ]; then max_count=$count best_gid_index="$idx" fi done # Use deterministic fallback if tied - prefer index 3 (SR-IOV standard) if [ ${#gid_index_count[@]} -gt 1 ]; then echo "[Infer RoCE] Multiple GID indices found, selecting most common: ${best_gid_index}" # If there's a tie, prefer index 3 as it's most common in SR-IOV setups if [ -n "${gid_index_count['3']}" ] && [ "${gid_index_count['3']}" -eq "$max_count" ]; then best_gid_index="3" echo "[Infer RoCE] Using deterministic fallback: GID_INDEX=3 (SR-IOV standard)" fi fi # Check if GID_INDEX is already set via environment variables if [ -n "${NCCL_IB_GID_INDEX}" ]; then echo "[Infer RoCE] Using pre-configured NCCL_IB_GID_INDEX=${NCCL_IB_GID_INDEX} from environment" export NVSHMEM_IB_GID_INDEX=${NVSHMEM_IB_GID_INDEX:-$NCCL_IB_GID_INDEX} export UCX_IB_GID_INDEX=${UCX_IB_GID_INDEX:-$NCCL_IB_GID_INDEX} echo "[Infer RoCE] Using pre-configured GID_INDEX=${NCCL_IB_GID_INDEX} for NCCL, NVSHMEM, and UCX" elif [ -n "$best_gid_index" ]; then echo "[Infer RoCE] Selected GID_INDEX: ${best_gid_index} (found on ${max_count} HCAs)" export NCCL_IB_GID_INDEX=${NCCL_IB_GID_INDEX:-$best_gid_index} export NVSHMEM_IB_GID_INDEX=${NVSHMEM_IB_GID_INDEX:-$best_gid_index} export UCX_IB_GID_INDEX=${UCX_IB_GID_INDEX:-$best_gid_index} echo "[Infer RoCE] Exported GID_INDEX=${best_gid_index} for NCCL, NVSHMEM, and UCX" else echo "[Infer RoCE] ERROR: No valid IPv4 ${KSERVE_INFER_IB_GID_INDEX_GREP} GID_INDEX found on any HCA." fi else echo "[Infer RoCE] No active HCAs found, skipping GID_INDEX inference." fi fi # --disable-access-log-for-endpoints landed in vLLM 0.16.0 (vllm-project/vllm#30011). # Older versions still need the blanket --disable-uvicorn-access-log. ACCESS_LOG_ARGS="--disable-uvicorn-access-log" VLLM_VERSION=$(vllm --version 2>/dev/null | tail -1 | awk '{print $NF}') echo "[access-log-detect] vllm version='${VLLM_VERSION}'" if [[ "$VLLM_VERSION" =~ ^[0-9]+\.[0-9]+ ]] && [ "$(printf '%s\n%s\n' "0.16.0" "${VLLM_VERSION}" | sort -V | head -1)" = "0.16.0" ]; then ACCESS_LOG_ARGS="--disable-access-log-for-endpoints /health,/metrics,/ping" fi echo "[access-log-detect] selected ACCESS_LOG_ARGS='${ACCESS_LOG_ARGS}'" # --shutdown-timeout landed in vLLM 0.18.0 (vllm-project/vllm#36666). SHUTDOWN_TIMEOUT_ARGS="" if [[ "$VLLM_VERSION" =~ ^[0-9]+\.[0-9]+ ]] && [ "$(printf '%s\n%s\n' "0.18.0" "${VLLM_VERSION}" | sort -V | head -1)" = "0.18.0" ]; then SHUTDOWN_TIMEOUT_ARGS="--shutdown-timeout 40" fi eval "exec vllm serve /mnt/models \ --served-model-name "facebook/opt-125m" "publishers/kserve-ci-e2e-test/models/facebook/opt-125m" \ --port 8001 \ ${ACCESS_LOG_ARGS} \ ${SHUTDOWN_TIMEOUT_ARGS} \ --enable-ssl-refresh \ --ssl-certfile /var/run/kserve/tls/tls.crt \ --ssl-keyfile /var/run/kserve/tls/tls.key \ ${VLLM_ADDITIONAL_ARGS} \ $@" - -- env: - name: HOME value: /home - name: VLLM_LOGGING_LEVEL value: DEBUG - name: VLLM_CPU_KVCACHE_SPACE value: "1" - name: VLLM_ENABLE_V1_MULTIPROCESSING value: "0" - name: USER value: nonroot - name: TORCHINDUCTOR_CACHE_DIR value: /tmp/torchinductor-cache - name: HF_HUB_CACHE value: /models image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imagePullPolicy: IfNotPresent lifecycle: preStop: exec: command: - /bin/sleep - "15" livenessProbe: failureThreshold: 8 httpGet: path: /health port: 8000 scheme: HTTPS initialDelaySeconds: 180 periodSeconds: 30 successThreshold: 1 timeoutSeconds: 30 name: main ports: - containerPort: 8001 protocol: TCP readinessProbe: failureThreshold: 3 httpGet: path: /health port: 8000 scheme: HTTPS initialDelaySeconds: 30 periodSeconds: 10 successThreshold: 1 timeoutSeconds: 5 resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL readOnlyRootFilesystem: true runAsNonRoot: true runAsUser: 1000690000 seccompProfile: type: RuntimeDefault startupProbe: failureThreshold: 60 httpGet: path: /health port: 8001 scheme: HTTPS periodSeconds: 10 successThreshold: 1 timeoutSeconds: 1 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /home name: home - mountPath: /tmp name: tmp-dir - mountPath: /dev/shm name: dshm - mountPath: /models name: model-cache - mountPath: /var/run/kserve/tls name: tls-certs readOnly: true - mountPath: /mnt/models name: kserve-pvc-source readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-cn767 readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-s8z92 - name: llmisvc-model-pvc-router-manage-e8706282-kserve-dockercfg-m8pgz initContainers: - command: - /app/pd-sidecar - --port=8000 - --vllm-port=8001 - --kv-connector=nixlv2 - --enable-ssrf-protection=true - --pool-group=inference.networking.x-k8s.io - --inference-pool=kserve-ci-e2e-test/llmisvc-model-pvc-router-manage-e8706282-inference-pool - --secure-proxy=true - --cert-path=/var/run/kserve/tls - --decoder-use-tls=true - --prefiller-use-tls=true env: - name: INFERENCE_POOL_NAMESPACE valueFrom: fieldRef: apiVersion: v1 fieldPath: metadata.namespace - name: SSL_CERT_DIR value: /var/run/kserve/tls:/var/run/secrets/kubernetes.io/serviceaccount:/etc/pki/tls/certs - name: INFERENCE_POOL_NAME value: llmisvc-model-pvc-router-manage-e8706282-inference-pool image: ghcr.io/llm-d/llm-d-router-disagg-sidecar:v0.9.0-rc.2 imagePullPolicy: IfNotPresent livenessProbe: failureThreshold: 3 httpGet: path: /health port: 8000 scheme: HTTPS initialDelaySeconds: 10 periodSeconds: 10 successThreshold: 1 timeoutSeconds: 10 name: llm-d-routing-sidecar ports: - containerPort: 8000 protocol: TCP readinessProbe: failureThreshold: 10 httpGet: path: /health port: 8000 scheme: HTTPS initialDelaySeconds: 10 periodSeconds: 10 successThreshold: 1 timeoutSeconds: 5 resources: {} restartPolicy: Always securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL readOnlyRootFilesystem: true runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /var/run/kserve/tls name: tls-certs readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-cn767 readOnly: true nodeName: ip-10-0-137-79.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: llmisvc-model-pvc-router-manage-e8706282-kserve serviceAccountName: llmisvc-model-pvc-router-manage-e8706282-kserve terminationGracePeriodSeconds: 60 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: - emptyDir: {} name: home - emptyDir: {} name: tmp-dir - emptyDir: medium: Memory sizeLimit: 1Gi name: dshm - emptyDir: {} name: model-cache - name: tls-certs secret: defaultMode: 420 secretName: llmisvb19f98874e050eec8ca94d49676113f0-kserve-self-signed-certs - name: kserve-pvc-source persistentVolumeClaim: claimName: e2e-pvc-model-storage - name: kube-api-access-cn767 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: "2026-07-09T18:46:57Z" observedGeneration: 1 status: "False" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-09T18:46:57Z" message: 'containers with incomplete status: [llm-d-routing-sidecar]' observedGeneration: 1 reason: ContainersNotInitialized status: "False" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-09T18:46:57Z" message: 'containers with unready status: [llm-d-routing-sidecar main]' observedGeneration: 1 reason: ContainersNotReady status: "False" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-09T18:46:57Z" message: 'containers with unready status: [llm-d-routing-sidecar main]' observedGeneration: 1 reason: ContainersNotReady status: "False" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-09T18:46:57Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: "" lastState: {} name: main ready: false restartCount: 0 started: false state: waiting: reason: PodInitializing volumeMounts: - mountPath: /home name: home - mountPath: /tmp name: tmp-dir - mountPath: /dev/shm name: dshm - mountPath: /models name: model-cache - mountPath: /var/run/kserve/tls name: tls-certs readOnly: true recursiveReadOnly: Disabled - mountPath: /mnt/models name: kserve-pvc-source readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-cn767 readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.137.79 hostIPs: - ip: 10.0.137.79 initContainerStatuses: - image: ghcr.io/llm-d/llm-d-router-disagg-sidecar:v0.9.0-rc.2 imageID: "" lastState: {} name: llm-d-routing-sidecar ready: false restartCount: 0 started: false state: waiting: reason: PodInitializing volumeMounts: - mountPath: /var/run/kserve/tls name: tls-certs readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-cn767 readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Pending qosClass: Burstable startTime: "2026-07-09T18:46:57Z" - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.132.0.52/23"],"mac_address":"0a:58:0a:84:00:34","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":"10.132.0.52/23","gateway_ip":"10.132.0.1","role":"primary"}}' openshift.io/scc: restricted-v2 seccomp.security.alpha.kubernetes.io/pod: runtime/default security.openshift.io/validated-scc-subject-type: user creationTimestamp: "2026-07-09T18:46:57Z" generateName: llmisvc-model-pvc-router-manage-e8706282-kserve-prefill-79497db4cc- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload-prefill app.kubernetes.io/name: llmisvc-model-pvc-router-manage-e8706282 app.kubernetes.io/part-of: llminferenceservice kserve.io/component: workload llm-d.ai/role: prefill pod-template-hash: 79497db4cc managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-137-79 operation: Update subresource: status time: "2026-07-09T18:46:57Z" - 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:kserve.io/component: {} f:llm-d.ai/role: {} f:pod-template-hash: {} f:ownerReferences: .: {} k:{"uid":"4062d307-81e2-4102-bc89-17ab94ee00ca"}: {} f:spec: f:containers: k:{"name":"main"}: .: {} f:command: {} f:env: .: {} k:{"name":"HF_HUB_CACHE"}: .: {} f:name: {} f:value: {} k:{"name":"HOME"}: .: {} f:name: {} f:value: {} k:{"name":"TORCHINDUCTOR_CACHE_DIR"}: .: {} f:name: {} f:value: {} k:{"name":"USER"}: .: {} f:name: {} f:value: {} k:{"name":"VLLM_CPU_KVCACHE_SPACE"}: .: {} f:name: {} f:value: {} k:{"name":"VLLM_ENABLE_V1_MULTIPROCESSING"}: .: {} f:name: {} f:value: {} k:{"name":"VLLM_LOGGING_LEVEL"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:lifecycle: .: {} f:preStop: .: {} f:exec: .: {} f:command: {} f:livenessProbe: .: {} f:failureThreshold: {} f:httpGet: .: {} f:path: {} f:port: {} f:scheme: {} f:initialDelaySeconds: {} f:periodSeconds: {} f:successThreshold: {} f:timeoutSeconds: {} f:name: {} f:ports: .: {} k:{"containerPort":8000,"protocol":"TCP"}: .: {} f:containerPort: {} f:protocol: {} f:readinessProbe: .: {} f:failureThreshold: {} f:httpGet: .: {} f:path: {} f:port: {} f:scheme: {} f:initialDelaySeconds: {} f:periodSeconds: {} f:successThreshold: {} f:timeoutSeconds: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} 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:periodSeconds: {} f:successThreshold: {} f:timeoutSeconds: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/dev/shm"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/home"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:readOnly: {} k:{"mountPath":"/models"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/tmp"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/kserve/tls"}: .: {} f:mountPath: {} f:name: {} f:readOnly: {} f:dnsPolicy: {} f:enableServiceLinks: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"dshm"}: .: {} f:emptyDir: .: {} f:medium: {} f:sizeLimit: {} f:name: {} k:{"name":"home"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"kserve-pvc-source"}: .: {} f:name: {} f:persistentVolumeClaim: .: {} f:claimName: {} k:{"name":"model-cache"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tls-certs"}: .: {} f:name: {} f:secret: .: {} f:defaultMode: {} f:secretName: {} k:{"name":"tmp-dir"}: .: {} f:emptyDir: {} f:name: {} manager: kube-controller-manager operation: Update time: "2026-07-09T18:46:57Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:status: f:conditions: k:{"type":"ContainersReady"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:message: {} f:observedGeneration: {} f:reason: {} f:status: {} f:type: {} k:{"type":"Initialized"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodReadyToStartContainers"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodScheduled"}: f:observedGeneration: {} k:{"type":"Ready"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:message: {} f:observedGeneration: {} f:reason: {} f:status: {} f:type: {} f:containerStatuses: {} f:hostIP: {} f:hostIPs: {} f:observedGeneration: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-09T18:46:57Z" name: llmisvc-model-pvc-router-manage-e8706282-kserve-prefill-7957mft namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: llmisvc-model-pvc-router-manage-e8706282-kserve-prefill-79497db4cc uid: 4062d307-81e2-4102-bc89-17ab94ee00ca resourceVersion: "47432" uid: e2fb5119-9261-41ca-8654-b7060a960639 spec: containers: - command: - /bin/bash - -c - |- if [ -f /etc/profile.d/ibm-aiu-setup.sh ]; then source /etc/profile.d/ibm-aiu-setup.sh fi if [ "$KSERVE_INFER_ROCE" = "true" ]; then echo "Trying to infer RoCE configs ... " grep -H . /sys/class/infiniband/*/ports/*/gids/* 2>/dev/null grep -H . /sys/class/infiniband/*/ports/*/gid_attrs/types/* 2>/dev/null cat /proc/driver/nvidia/params KSERVE_INFER_IB_GID_INDEX_GREP=${KSERVE_INFER_IB_GID_INDEX_GREP:-"RoCE v2"} echo "[Infer RoCE] Discovering active HCAs ..." active_hcas=() # Loop through all mlx5 devices found in sysfs for hca_dir in /sys/class/infiniband/mlx5_*; do # Ensure it's a directory before proceeding if [ -d "$hca_dir" ]; then hca_name=$(basename "$hca_dir") port_state_file="$hca_dir/ports/1/state" # Assume port 1 type_file="$hca_dir/ports/1/gid_attrs/types/*" echo "[Infer RoCE] Check if the port state file ${port_state_file} exists and contains 'ACTIVE'" if [ -f "$port_state_file" ] && grep -q "ACTIVE" "$port_state_file" && grep -q "${KSERVE_INFER_IB_GID_INDEX_GREP}" ${type_file} 2>/dev/null; then echo "[Infer RoCE] Found active HCA: $hca_name" active_hcas+=("$hca_name") else echo "[Infer RoCE] Skipping inactive or down HCA: $hca_name" fi fi done # Check if we found any active HCAs if [ ${#active_hcas[@]} -gt 0 ]; then # Join the array elements with a comma hca_port_pairs=() for hca in "${active_hcas[@]}"; do hca_port_pairs+=("${hca}:1") done active_hca_list=$(IFS=,; echo "${active_hcas[*]}") hca_port_pairs_list=$(IFS=,; echo "${hca_port_pairs[*]}") echo "[Infer RoCE] Setting active HCAs: ${active_hca_list}" export NCCL_IB_HCA=${NCCL_IB_HCA:-${active_hca_list}} export NVSHMEM_HCA_LIST=${NVSHMEM_HCA_LIST:-${hca_port_pairs_list}} export UCX_NET_DEVICES=${UCX_NET_DEVICES:-${hca_port_pairs_list}} echo "[Infer RoCE] NCCL_IB_HCA=${NCCL_IB_HCA}" echo "[Infer RoCE] NVSHMEM_HCA_LIST=${NVSHMEM_HCA_LIST}" echo "[Infer RoCE] UCX_NET_DEVICES=${UCX_NET_DEVICES}" else echo "[Infer RoCE] WARNING: No active RoCE HCAs found. NCCL_IB_HCA will not be set." fi if [ ${#active_hcas[@]} -gt 0 ]; then echo "[Infer RoCE] Finding GID_INDEX for each active HCA (SR-IOV compatible)..." # For SR-IOV environments, find the most common IPv4 RoCE v2 GID index across all HCAs declare -A gid_index_count declare -A hca_gid_index for hca_name in "${active_hcas[@]}"; do echo "[Infer RoCE] Processing HCA: ${hca_name}" # Find all RoCE v2 IPv4 GIDs for this HCA and count by index for tpath in /sys/class/infiniband/${hca_name}/ports/1/gid_attrs/types/*; do if grep -q "${KSERVE_INFER_IB_GID_INDEX_GREP}" "$tpath" 2>/dev/null; then idx=$(basename "$tpath") gid_file="/sys/class/infiniband/${hca_name}/ports/1/gids/${idx}" # Check for IPv4 GID (contains ffff:) if [ -f "$gid_file" ] && grep -q "ffff:" "$gid_file"; then gid_value=$(cat "$gid_file" 2>/dev/null || echo "") echo "[Infer RoCE] Found IPv4 RoCE v2 GID for ${hca_name}: index=${idx}, gid=${gid_value}" hca_gid_index["${hca_name}"]="${idx}" gid_index_count["${idx}"]=$((${gid_index_count["${idx}"]} + 1)) break # Use first found IPv4 GID per HCA fi fi done done # Find the most common GID index (most likely to be consistent across nodes) best_gid_index="" max_count=0 for idx in "${!gid_index_count[@]}"; do count=${gid_index_count["${idx}"]} echo "[Infer RoCE] GID_INDEX ${idx} found on ${count} HCAs" if [ $count -gt $max_count ]; then max_count=$count best_gid_index="$idx" fi done # Use deterministic fallback if tied - prefer index 3 (SR-IOV standard) if [ ${#gid_index_count[@]} -gt 1 ]; then echo "[Infer RoCE] Multiple GID indices found, selecting most common: ${best_gid_index}" # If there's a tie, prefer index 3 as it's most common in SR-IOV setups if [ -n "${gid_index_count['3']}" ] && [ "${gid_index_count['3']}" -eq "$max_count" ]; then best_gid_index="3" echo "[Infer RoCE] Using deterministic fallback: GID_INDEX=3 (SR-IOV standard)" fi fi # Check if GID_INDEX is already set via environment variables if [ -n "${NCCL_IB_GID_INDEX}" ]; then echo "[Infer RoCE] Using pre-configured NCCL_IB_GID_INDEX=${NCCL_IB_GID_INDEX} from environment" export NVSHMEM_IB_GID_INDEX=${NVSHMEM_IB_GID_INDEX:-$NCCL_IB_GID_INDEX} export UCX_IB_GID_INDEX=${UCX_IB_GID_INDEX:-$NCCL_IB_GID_INDEX} echo "[Infer RoCE] Using pre-configured GID_INDEX=${NCCL_IB_GID_INDEX} for NCCL, NVSHMEM, and UCX" elif [ -n "$best_gid_index" ]; then echo "[Infer RoCE] Selected GID_INDEX: ${best_gid_index} (found on ${max_count} HCAs)" export NCCL_IB_GID_INDEX=${NCCL_IB_GID_INDEX:-$best_gid_index} export NVSHMEM_IB_GID_INDEX=${NVSHMEM_IB_GID_INDEX:-$best_gid_index} export UCX_IB_GID_INDEX=${UCX_IB_GID_INDEX:-$best_gid_index} echo "[Infer RoCE] Exported GID_INDEX=${best_gid_index} for NCCL, NVSHMEM, and UCX" else echo "[Infer RoCE] ERROR: No valid IPv4 ${KSERVE_INFER_IB_GID_INDEX_GREP} GID_INDEX found on any HCA." fi else echo "[Infer RoCE] No active HCAs found, skipping GID_INDEX inference." fi fi # --disable-access-log-for-endpoints landed in vLLM 0.16.0 (vllm-project/vllm#30011). # Older versions still need the blanket --disable-uvicorn-access-log. ACCESS_LOG_ARGS="--disable-uvicorn-access-log" VLLM_VERSION=$(vllm --version 2>/dev/null | tail -1 | awk '{print $NF}') echo "[access-log-detect] vllm version='${VLLM_VERSION}'" if [[ "$VLLM_VERSION" =~ ^[0-9]+\.[0-9]+ ]] && [ "$(printf '%s\n%s\n' "0.16.0" "${VLLM_VERSION}" | sort -V | head -1)" = "0.16.0" ]; then ACCESS_LOG_ARGS="--disable-access-log-for-endpoints /health,/metrics,/ping" fi echo "[access-log-detect] selected ACCESS_LOG_ARGS='${ACCESS_LOG_ARGS}'" # --shutdown-timeout landed in vLLM 0.18.0 (vllm-project/vllm#36666). SHUTDOWN_TIMEOUT_ARGS="" if [[ "$VLLM_VERSION" =~ ^[0-9]+\.[0-9]+ ]] && [ "$(printf '%s\n%s\n' "0.18.0" "${VLLM_VERSION}" | sort -V | head -1)" = "0.18.0" ]; then SHUTDOWN_TIMEOUT_ARGS="--shutdown-timeout 40" fi eval "exec vllm serve /mnt/models \ --served-model-name "facebook/opt-125m" \ --port 8000 \ ${ACCESS_LOG_ARGS} \ ${SHUTDOWN_TIMEOUT_ARGS} \ --enable-ssl-refresh \ --ssl-certfile /var/run/kserve/tls/tls.crt \ --ssl-keyfile /var/run/kserve/tls/tls.key \ ${VLLM_ADDITIONAL_ARGS} \ $@" - -- env: - name: HOME value: /home - name: VLLM_LOGGING_LEVEL value: DEBUG - name: VLLM_CPU_KVCACHE_SPACE value: "1" - name: VLLM_ENABLE_V1_MULTIPROCESSING value: "0" - name: USER value: nonroot - name: TORCHINDUCTOR_CACHE_DIR value: /tmp/torchinductor-cache - name: HF_HUB_CACHE value: /models image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imagePullPolicy: IfNotPresent lifecycle: preStop: exec: command: - /bin/sleep - "15" livenessProbe: failureThreshold: 8 httpGet: path: /health port: 8000 scheme: HTTPS initialDelaySeconds: 180 periodSeconds: 30 successThreshold: 1 timeoutSeconds: 30 name: main ports: - containerPort: 8000 protocol: TCP readinessProbe: failureThreshold: 3 httpGet: path: /health port: 8000 scheme: HTTPS initialDelaySeconds: 30 periodSeconds: 10 successThreshold: 1 timeoutSeconds: 5 resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL readOnlyRootFilesystem: true runAsNonRoot: true runAsUser: 1000690000 seccompProfile: type: RuntimeDefault startupProbe: failureThreshold: 60 httpGet: path: /health port: 8000 scheme: HTTPS periodSeconds: 10 successThreshold: 1 timeoutSeconds: 1 terminationMessagePath: /dev/termination-log terminationMessagePolicy: File volumeMounts: - mountPath: /home name: home - mountPath: /tmp name: tmp-dir - mountPath: /dev/shm name: dshm - mountPath: /models name: model-cache - mountPath: /var/run/kserve/tls name: tls-certs readOnly: true - mountPath: /mnt/models name: kserve-pvc-source readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-4jrzq readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-s8z92 nodeName: ip-10-0-137-79.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: default serviceAccountName: default terminationGracePeriodSeconds: 60 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: - emptyDir: {} name: home - emptyDir: {} name: tmp-dir - emptyDir: medium: Memory sizeLimit: 1Gi name: dshm - emptyDir: {} name: model-cache - name: tls-certs secret: defaultMode: 420 secretName: llmisvb19f98874e050eec8ca94d49676113f0-kserve-self-signed-certs - name: kserve-pvc-source persistentVolumeClaim: claimName: e2e-pvc-model-storage - name: kube-api-access-4jrzq 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: "2026-07-09T18:46:57Z" observedGeneration: 1 status: "False" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-09T18:46:57Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-09T18:46:57Z" message: 'containers with unready status: [main]' observedGeneration: 1 reason: ContainersNotReady status: "False" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-09T18:46:57Z" message: 'containers with unready status: [main]' observedGeneration: 1 reason: ContainersNotReady status: "False" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-09T18:46:57Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: "" lastState: {} name: main ready: false restartCount: 0 started: false state: waiting: reason: ContainerCreating volumeMounts: - mountPath: /home name: home - mountPath: /tmp name: tmp-dir - mountPath: /dev/shm name: dshm - mountPath: /models name: model-cache - mountPath: /var/run/kserve/tls name: tls-certs readOnly: true recursiveReadOnly: Disabled - mountPath: /mnt/models name: kserve-pvc-source readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-4jrzq readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.137.79 hostIPs: - ip: 10.0.137.79 observedGeneration: 1 phase: Pending qosClass: Burstable startTime: "2026-07-09T18:46:57Z" - apiVersion: v1 kind: Pod metadata: annotations: app.kubernetes.io/version: 0.9.0 certificates.kserve.io/expiration-v2: "true" k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.134.0.44/23"],"mac_address":"0a:58:0a:86:00:2c","gateway_ips":["10.134.0.1"],"routes":[{"dest":"10.132.0.0/14","nextHop":"10.134.0.1"},{"dest":"172.31.0.0/16","nextHop":"10.134.0.1"},{"dest":"169.254.0.5/32","nextHop":"10.134.0.1"},{"dest":"100.64.0.0/16","nextHop":"10.134.0.1"}],"ip_address":"10.134.0.44/23","gateway_ip":"10.134.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.134.0.44" ], "mac": "0a:58:0a:86:00:2c", "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: "2026-07-09T18:31:36Z" generateName: llmisvc8bf079eb6eda4debfb4ef5bb7817824c-kserve-router-scheduler-5f555d4d85- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-router-scheduler app.kubernetes.io/name: llmisvc-model-pvc-router-manage-59b9d263 app.kubernetes.io/part-of: llminferenceservice pod-template-hash: 5f555d4d85 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-129-87 operation: Update subresource: status time: "2026-07-09T18:31:36Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: .: {} f:app.kubernetes.io/version: {} f:certificates.kserve.io/expiration-v2: {} 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":"f5034118-b204-4769-bbe1-12cbfea939d0"}: {} 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:lifecycle: .: {} f:preStop: .: {} f:exec: .: {} f:command: {} 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: {} f:dnsPolicy: {} f:enableServiceLinks: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"kserve-pvc-source"}: .: {} f:name: {} f:persistentVolumeClaim: .: {} f:claimName: {} 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: "2026-07-09T18:31:36Z" - 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: "2026-07-09T18:31:37Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:status: f:conditions: k:{"type":"ContainersReady"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"Initialized"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodReadyToStartContainers"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodScheduled"}: f:observedGeneration: {} k:{"type":"Ready"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} f:containerStatuses: {} f:hostIP: {} f:hostIPs: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.134.0.44"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-09T18:32:08Z" name: llmisvc8bf079eb6eda4debfb4ef5bb7817824c-kserve-router-schebclx5 namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: llmisvc8bf079eb6eda4debfb4ef5bb7817824c-kserve-router-scheduler-5f555d4d85 uid: f5034118-b204-4769-bbe1-12cbfea939d0 resourceVersion: "36413" uid: 0b12fcf9-7192-4135-a99e-9d03a833fd7b 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 - llmisvc-model-pvc-router-manage-59b9d263-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-router-endpoint-picker:v0.9.0-rc.2 imagePullPolicy: IfNotPresent lifecycle: preStop: exec: command: - /bin/sleep - "15" 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-f64bm readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-s8z92 - name: llmisvc-model-pvc-router-manage-59b9d263-epp-sa-dockercfg-rrhx2 nodeName: ip-10-0-129-87.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: llmisvc-model-pvc-router-manage-59b9d263-epp-sa serviceAccountName: llmisvc-model-pvc-router-manage-59b9d263-epp-sa terminationGracePeriodSeconds: 60 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: llmisv8bf079eb6eda4debfb4ef5bb7817824c-kserve-self-signed-certs - emptyDir: {} name: tokenizer-uds - emptyDir: {} name: tokenizer-tmp - emptyDir: {} name: tokenizer-cache - name: kserve-pvc-source persistentVolumeClaim: claimName: e2e-pvc-model-storage - name: kube-api-access-f64bm 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: "2026-07-09T18:31:37Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-09T18:31:36Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-09T18:32:08Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-09T18:32:08Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-09T18:31:36Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 256m memory: 500Mi containerID: cri-o://150dc5eb845cd9ec2c4a8add38fe6e3850be4c745b09a94bc1be33266e7521bf image: ghcr.io/llm-d/llm-d-router-endpoint-picker:v0.9.0-rc.2 imageID: ghcr.io/llm-d/llm-d-router-endpoint-picker@sha256:06b6c75d77afd0e07053402752a9736c2dfbc12a306d0d37d963aac4c1d4e6a6 lastState: {} name: main ready: true resources: requests: cpu: 256m memory: 500Mi restartCount: 0 started: true state: running: startedAt: "2026-07-09T18:31:37Z" 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-f64bm readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.129.87 hostIPs: - ip: 10.0.129.87 observedGeneration: 1 phase: Running podIP: 10.134.0.44 podIPs: - ip: 10.134.0.44 qosClass: Burstable startTime: "2026-07-09T18:31:36Z" - apiVersion: v1 kind: Pod metadata: annotations: app.kubernetes.io/version: 0.9.0 certificates.kserve.io/expiration-v2: "true" k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.133.0.51/23"],"mac_address":"0a:58:0a:85:00:33","gateway_ips":["10.133.0.1"],"routes":[{"dest":"10.132.0.0/14","nextHop":"10.133.0.1"},{"dest":"172.31.0.0/16","nextHop":"10.133.0.1"},{"dest":"169.254.0.5/32","nextHop":"10.133.0.1"},{"dest":"100.64.0.0/16","nextHop":"10.133.0.1"}],"ip_address":"10.133.0.51/23","gateway_ip":"10.133.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.133.0.51" ], "mac": "0a:58:0a:85:00:33", "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: "2026-07-09T18:46:57Z" generateName: llmisvcb19f98874e050eec8ca94d49676113f0-kserve-router-scheduler-67b4bb9646- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-router-scheduler app.kubernetes.io/name: llmisvc-model-pvc-router-manage-e8706282 app.kubernetes.io/part-of: llminferenceservice pod-template-hash: 67b4bb9646 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-131-205 operation: Update subresource: status time: "2026-07-09T18:46:57Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: .: {} f:app.kubernetes.io/version: {} f:certificates.kserve.io/expiration-v2: {} 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":"756f70d8-4dd7-41e6-8927-8fa13f774ffe"}: {} 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:lifecycle: .: {} f:preStop: .: {} f:exec: .: {} f:command: {} 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: {} f:dnsPolicy: {} f:enableServiceLinks: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"kserve-pvc-source"}: .: {} f:name: {} f:persistentVolumeClaim: .: {} f:claimName: {} 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: "2026-07-09T18:46:57Z" - 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: "2026-07-09T18:46:58Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:status: f:conditions: k:{"type":"ContainersReady"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"Initialized"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodReadyToStartContainers"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodScheduled"}: f:observedGeneration: {} k:{"type":"Ready"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} f:containerStatuses: {} f:hostIP: {} f:hostIPs: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.133.0.51"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-09T18:47:29Z" name: llmisvcb19f98874e050eec8ca94d49676113f0-kserve-router-sche9hm9n namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: llmisvcb19f98874e050eec8ca94d49676113f0-kserve-router-scheduler-67b4bb9646 uid: 756f70d8-4dd7-41e6-8927-8fa13f774ffe resourceVersion: "47942" uid: 39279130-b4d1-4a37-937a-6d88e93dcbaf spec: containers: - args: - --config-text - | apiVersion: inference.networking.x-k8s.io/v1alpha1 kind: EndpointPickerConfig plugins: - type: disagg-headers-handler - type: prefill-filter - type: decode-filter - type: queue-scorer - type: prefix-cache-scorer - type: max-score-picker - type: always-disagg-pd-decider - parameters: deciders: prefill: always-disagg-pd-decider type: disagg-profile-handler schedulingProfiles: - name: prefill plugins: - pluginRef: prefill-filter - pluginRef: queue-scorer weight: 2 - pluginRef: prefix-cache-scorer weight: 3 - pluginRef: max-score-picker - name: decode plugins: - pluginRef: decode-filter - pluginRef: queue-scorer weight: 2 - pluginRef: prefix-cache-scorer weight: 3 - pluginRef: max-score-picker command: - /app/epp - --pool-name - llmisvc-model-pvc-router-manage-e8706282-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-router-endpoint-picker:v0.9.0-rc.2 imagePullPolicy: IfNotPresent lifecycle: preStop: exec: command: - /bin/sleep - "15" 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-dv6md readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-s8z92 - name: llmisvc-model-pvc-router-manage-e8706282-epp-sa-dockercfg-ggvbd nodeName: ip-10-0-131-205.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: llmisvc-model-pvc-router-manage-e8706282-epp-sa serviceAccountName: llmisvc-model-pvc-router-manage-e8706282-epp-sa terminationGracePeriodSeconds: 60 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: llmisvb19f98874e050eec8ca94d49676113f0-kserve-self-signed-certs - emptyDir: {} name: tokenizer-uds - emptyDir: {} name: tokenizer-tmp - emptyDir: {} name: tokenizer-cache - name: kserve-pvc-source persistentVolumeClaim: claimName: e2e-pvc-model-storage - name: kube-api-access-dv6md 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: "2026-07-09T18:46:58Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-09T18:46:57Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-09T18:47:29Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-09T18:47:29Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-09T18:46:57Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 256m memory: 500Mi containerID: cri-o://4d158e154f7002ccae8dabc8d99437414cf50d3b5417d9fec1c0ba5b81119159 image: ghcr.io/llm-d/llm-d-router-endpoint-picker:v0.9.0-rc.2 imageID: ghcr.io/llm-d/llm-d-router-endpoint-picker@sha256:06b6c75d77afd0e07053402752a9736c2dfbc12a306d0d37d963aac4c1d4e6a6 lastState: {} name: main ready: true resources: requests: cpu: 256m memory: 500Mi restartCount: 0 started: true state: running: startedAt: "2026-07-09T18:46:58Z" 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-dv6md readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.131.205 hostIPs: - ip: 10.0.131.205 observedGeneration: 1 phase: Running podIP: 10.133.0.51 podIPs: - ip: 10.133.0.51 qosClass: Burstable startTime: "2026-07-09T18:46:57Z" - apiVersion: v1 kind: Pod metadata: annotations: istio.io/rev: openshift-gateway k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.134.0.45/23"],"mac_address":"0a:58:0a:86:00:2d","gateway_ips":["10.134.0.1"],"routes":[{"dest":"10.132.0.0/14","nextHop":"10.134.0.1"},{"dest":"172.31.0.0/16","nextHop":"10.134.0.1"},{"dest":"169.254.0.5/32","nextHop":"10.134.0.1"},{"dest":"100.64.0.0/16","nextHop":"10.134.0.1"}],"ip_address":"10.134.0.45/23","gateway_ip":"10.134.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.134.0.45" ], "mac": "0a:58:0a:86:00:2d", "default": true, "dns": {} }] openshift.io/scc: restricted-v2 prometheus.io/path: /stats/prometheus prometheus.io/port: "15020" prometheus.io/scrape: "true" seccomp.security.alpha.kubernetes.io/pod: runtime/default security.openshift.io/validated-scc-subject-type: user creationTimestamp: "2026-07-09T18:34:12Z" generateName: router-gateway-1-openshift-default-54c47cdfc7- generation: 1 labels: gateway.istio.io/managed: istio.io-gateway-controller gateway.networking.k8s.io/gateway-name: router-gateway-1 pod-template-hash: 54c47cdfc7 service.istio.io/canonical-name: router-gateway-1-openshift-default service.istio.io/canonical-revision: latest sidecar.istio.io/inject: "false" managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-129-87 operation: Update subresource: status time: "2026-07-09T18:34:12Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: .: {} f:istio.io/rev: {} f:prometheus.io/path: {} f:prometheus.io/port: {} f:prometheus.io/scrape: {} f:generateName: {} f:labels: .: {} f:gateway.istio.io/managed: {} f:gateway.networking.k8s.io/gateway-name: {} f:pod-template-hash: {} f:service.istio.io/canonical-name: {} f:service.istio.io/canonical-revision: {} f:sidecar.istio.io/inject: {} f:ownerReferences: .: {} k:{"uid":"19cae161-c64c-4946-8e69-21284ec458ab"}: {} f:spec: f:containers: k:{"name":"istio-proxy"}: .: {} f:args: {} f:env: .: {} k:{"name":"CA_ADDR"}: .: {} f:name: {} f:value: {} k:{"name":"GOMAXPROCS"}: .: {} f:name: {} f:valueFrom: .: {} f:resourceFieldRef: {} k:{"name":"GOMEMLIMIT"}: .: {} f:name: {} f:valueFrom: .: {} f:resourceFieldRef: {} k:{"name":"HOST_IP"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"INSTANCE_IP"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"ISTIO_CPU_LIMIT"}: .: {} f:name: {} f:valueFrom: .: {} f:resourceFieldRef: {} k:{"name":"ISTIO_META_APP_CONTAINERS"}: .: {} f:name: {} k:{"name":"ISTIO_META_CLUSTER_ID"}: .: {} f:name: {} f:value: {} k:{"name":"ISTIO_META_INTERCEPTION_MODE"}: .: {} f:name: {} f:value: {} k:{"name":"ISTIO_META_MESH_ID"}: .: {} f:name: {} f:value: {} k:{"name":"ISTIO_META_NODE_NAME"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"ISTIO_META_OWNER"}: .: {} f:name: {} f:value: {} k:{"name":"ISTIO_META_POD_PORTS"}: .: {} f:name: {} f:value: {} k:{"name":"ISTIO_META_WORKLOAD_NAME"}: .: {} f:name: {} f:value: {} k:{"name":"PILOT_CERT_PROVIDER"}: .: {} f:name: {} f:value: {} k:{"name":"POD_NAME"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"POD_NAMESPACE"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"PROXY_CONFIG"}: .: {} f:name: {} f:value: {} k:{"name":"SERVICE_ACCOUNT"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"TRUST_DOMAIN"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:ports: .: {} k:{"containerPort":15020,"protocol":"TCP"}: .: {} f:containerPort: {} f:name: {} f:protocol: {} k:{"containerPort":15021,"protocol":"TCP"}: .: {} f:containerPort: {} f:name: {} f:protocol: {} k:{"containerPort":15090,"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:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:securityContext: .: {} f:allowPrivilegeEscalation: {} f:capabilities: .: {} f:drop: {} f:privileged: {} f:readOnlyRootFilesystem: {} f:runAsGroup: {} f:runAsNonRoot: {} f:runAsUser: {} 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":"/etc/istio/pod"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/etc/istio/proxy"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/lib/istio/data"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/secrets/credential-uds"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/secrets/istio"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/secrets/tokens"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/secrets/workload-spiffe-credentials"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/secrets/workload-spiffe-uds"}: .: {} f:mountPath: {} f:name: {} f:dnsPolicy: {} f:enableServiceLinks: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: .: {} f:sysctls: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"credential-socket"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"istio-data"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"istio-envoy"}: .: {} f:emptyDir: .: {} f:medium: {} f:name: {} k:{"name":"istio-podinfo"}: .: {} f:downwardAPI: .: {} f:defaultMode: {} f:items: {} f:name: {} k:{"name":"istio-token"}: .: {} f:name: {} f:projected: .: {} f:defaultMode: {} f:sources: {} k:{"name":"istiod-ca-cert"}: .: {} f:configMap: .: {} f:defaultMode: {} f:name: {} f:name: {} k:{"name":"workload-certs"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"workload-socket"}: .: {} f:emptyDir: {} f:name: {} manager: kube-controller-manager operation: Update time: "2026-07-09T18:34:12Z" - 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: "2026-07-09T18:34:13Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:status: f:conditions: k:{"type":"ContainersReady"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"Initialized"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodReadyToStartContainers"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodScheduled"}: f:observedGeneration: {} k:{"type":"Ready"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} f:containerStatuses: {} f:hostIP: {} f:hostIPs: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.134.0.45"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-09T18:34:30Z" name: router-gateway-1-openshift-default-54c47cdfc7-wl8wv namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: router-gateway-1-openshift-default-54c47cdfc7 uid: 19cae161-c64c-4946-8e69-21284ec458ab resourceVersion: "38210" uid: de7a8044-a4f2-47c6-bbdc-04fec9ff8da4 spec: containers: - args: - proxy - router - --domain - $(POD_NAMESPACE).svc.cluster.local - --proxyLogLevel - warning - --proxyComponentLogLevel - misc:error - --log_output_level - default:info env: - name: PILOT_CERT_PROVIDER value: istiod - name: CA_ADDR value: istiod-openshift-gateway.openshift-ingress.svc:15012 - name: POD_NAME valueFrom: fieldRef: apiVersion: v1 fieldPath: metadata.name - name: POD_NAMESPACE valueFrom: fieldRef: apiVersion: v1 fieldPath: metadata.namespace - name: INSTANCE_IP valueFrom: fieldRef: apiVersion: v1 fieldPath: status.podIP - name: SERVICE_ACCOUNT valueFrom: fieldRef: apiVersion: v1 fieldPath: spec.serviceAccountName - name: HOST_IP valueFrom: fieldRef: apiVersion: v1 fieldPath: status.hostIP - name: ISTIO_CPU_LIMIT valueFrom: resourceFieldRef: divisor: "0" resource: limits.cpu - name: PROXY_CONFIG value: | {"discoveryAddress":"istiod-openshift-gateway.openshift-ingress.svc:15012","proxyHeaders":{"server":{"disabled":true},"envoyDebugHeaders":{"disabled":true},"metadataExchangeHeaders":{"mode":"IN_MESH"}}} - name: ISTIO_META_POD_PORTS value: '[]' - name: ISTIO_META_APP_CONTAINERS - name: GOMEMLIMIT valueFrom: resourceFieldRef: divisor: "0" resource: limits.memory - name: GOMAXPROCS valueFrom: resourceFieldRef: divisor: "0" resource: limits.cpu - name: ISTIO_META_CLUSTER_ID value: Kubernetes - name: ISTIO_META_NODE_NAME valueFrom: fieldRef: apiVersion: v1 fieldPath: spec.nodeName - name: ISTIO_META_INTERCEPTION_MODE value: REDIRECT - name: ISTIO_META_WORKLOAD_NAME value: router-gateway-1-openshift-default - name: ISTIO_META_OWNER value: kubernetes://apis/apps/v1/namespaces/kserve-ci-e2e-test/deployments/router-gateway-1-openshift-default - name: ISTIO_META_MESH_ID value: cluster.local - name: TRUST_DOMAIN value: cluster.local image: registry.redhat.io/openshift-service-mesh/istio-proxyv2-rhel9@sha256:7d15cebf9b62f3f235c0eab5158ac8ff2fda86a1d193490dc94c301402c99da8 imagePullPolicy: IfNotPresent name: istio-proxy ports: - containerPort: 15020 name: metrics protocol: TCP - containerPort: 15021 name: status-port protocol: TCP - containerPort: 15090 name: http-envoy-prom protocol: TCP readinessProbe: failureThreshold: 4 httpGet: path: /healthz/ready port: 15021 scheme: HTTP periodSeconds: 15 successThreshold: 1 timeoutSeconds: 1 resources: limits: cpu: "2" memory: 2Gi requests: cpu: 100m memory: 256Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL privileged: false readOnlyRootFilesystem: true runAsGroup: 1000699999 runAsNonRoot: true runAsUser: 1000699999 startupProbe: failureThreshold: 30 httpGet: path: /healthz/ready port: 15021 scheme: HTTP initialDelaySeconds: 1 periodSeconds: 1 successThreshold: 1 timeoutSeconds: 1 terminationMessagePath: /dev/termination-log terminationMessagePolicy: File volumeMounts: - mountPath: /var/run/secrets/workload-spiffe-uds name: workload-socket - mountPath: /var/run/secrets/credential-uds name: credential-socket - mountPath: /var/run/secrets/workload-spiffe-credentials name: workload-certs - mountPath: /var/run/secrets/istio name: istiod-ca-cert - mountPath: /var/lib/istio/data name: istio-data - mountPath: /etc/istio/proxy name: istio-envoy - mountPath: /var/run/secrets/tokens name: istio-token - mountPath: /etc/istio/pod name: istio-podinfo - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-rrh7h readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: router-gateway-1-openshift-default-dockercfg-7sv49 nodeName: ip-10-0-129-87.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault sysctls: - name: net.ipv4.ip_unprivileged_port_start value: "0" serviceAccount: router-gateway-1-openshift-default serviceAccountName: router-gateway-1-openshift-default 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: - emptyDir: {} name: workload-socket - emptyDir: {} name: credential-socket - emptyDir: {} name: workload-certs - emptyDir: medium: Memory name: istio-envoy - emptyDir: {} name: istio-data - downwardAPI: defaultMode: 420 items: - fieldRef: apiVersion: v1 fieldPath: metadata.labels path: labels - fieldRef: apiVersion: v1 fieldPath: metadata.annotations path: annotations name: istio-podinfo - name: istio-token projected: defaultMode: 420 sources: - serviceAccountToken: audience: istio-ca expirationSeconds: 43200 path: istio-token - configMap: defaultMode: 420 name: openshift-gw-ca-root-cert name: istiod-ca-cert - name: kube-api-access-rrh7h 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: "2026-07-09T18:34:13Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-09T18:34:12Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-09T18:34:30Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-09T18:34:30Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-09T18:34:12Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 100m memory: 256Mi containerID: cri-o://c88dc4b859ceaefb7f43f51cbed5605cfac890243c9d752a17eec785df43129a image: registry.redhat.io/openshift-service-mesh/istio-proxyv2-rhel9@sha256:7d15cebf9b62f3f235c0eab5158ac8ff2fda86a1d193490dc94c301402c99da8 imageID: registry.redhat.io/openshift-service-mesh/istio-proxyv2-rhel9@sha256:0ae6667459cd9e9ea0549e27bb014c722924454bf594b31646df8df58be91cc5 lastState: terminated: containerID: cri-o://a3f047eea2d4b9769553d4c5676282f6d164c49022ab69481368f20e3c481c38 exitCode: 0 finishedAt: "2026-07-09T18:34:23Z" reason: Completed startedAt: "2026-07-09T18:34:13Z" name: istio-proxy ready: true resources: limits: cpu: "2" memory: 2Gi requests: cpu: 100m memory: 256Mi restartCount: 1 started: true state: running: startedAt: "2026-07-09T18:34:23Z" user: linux: gid: 1000699999 supplementalGroups: - 1000699999 - 1000690000 uid: 1000699999 volumeMounts: - mountPath: /var/run/secrets/workload-spiffe-uds name: workload-socket - mountPath: /var/run/secrets/credential-uds name: credential-socket - mountPath: /var/run/secrets/workload-spiffe-credentials name: workload-certs - mountPath: /var/run/secrets/istio name: istiod-ca-cert - mountPath: /var/lib/istio/data name: istio-data - mountPath: /etc/istio/proxy name: istio-envoy - mountPath: /var/run/secrets/tokens name: istio-token - mountPath: /etc/istio/pod name: istio-podinfo - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-rrh7h readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.129.87 hostIPs: - ip: 10.0.129.87 observedGeneration: 1 phase: Running podIP: 10.134.0.45 podIPs: - ip: 10.134.0.45 qosClass: Burstable startTime: "2026-07-09T18:34:12Z" - apiVersion: v1 kind: Pod metadata: annotations: istio.io/rev: openshift-gateway k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.132.0.48/23"],"mac_address":"0a:58:0a:84:00:30","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":"10.132.0.48/23","gateway_ip":"10.132.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.132.0.48" ], "mac": "0a:58:0a:84:00:30", "default": true, "dns": {} }] openshift.io/scc: restricted-v2 prometheus.io/path: /stats/prometheus prometheus.io/port: "15020" prometheus.io/scrape: "true" seccomp.security.alpha.kubernetes.io/pod: runtime/default security.openshift.io/validated-scc-subject-type: user creationTimestamp: "2026-07-09T18:44:51Z" generateName: router-gateway-2-openshift-default-684dc9b4b9- generation: 1 labels: gateway.istio.io/managed: istio.io-gateway-controller gateway.networking.k8s.io/gateway-name: router-gateway-2 pod-template-hash: 684dc9b4b9 service.istio.io/canonical-name: router-gateway-2-openshift-default service.istio.io/canonical-revision: latest sidecar.istio.io/inject: "false" managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-137-79 operation: Update subresource: status time: "2026-07-09T18:44:51Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: .: {} f:istio.io/rev: {} f:prometheus.io/path: {} f:prometheus.io/port: {} f:prometheus.io/scrape: {} f:generateName: {} f:labels: .: {} f:gateway.istio.io/managed: {} f:gateway.networking.k8s.io/gateway-name: {} f:pod-template-hash: {} f:service.istio.io/canonical-name: {} f:service.istio.io/canonical-revision: {} f:sidecar.istio.io/inject: {} f:ownerReferences: .: {} k:{"uid":"5987bc37-5e76-466c-9d39-5ea3b765ad9e"}: {} f:spec: f:containers: k:{"name":"istio-proxy"}: .: {} f:args: {} f:env: .: {} k:{"name":"CA_ADDR"}: .: {} f:name: {} f:value: {} k:{"name":"GOMAXPROCS"}: .: {} f:name: {} f:valueFrom: .: {} f:resourceFieldRef: {} k:{"name":"GOMEMLIMIT"}: .: {} f:name: {} f:valueFrom: .: {} f:resourceFieldRef: {} k:{"name":"HOST_IP"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"INSTANCE_IP"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"ISTIO_CPU_LIMIT"}: .: {} f:name: {} f:valueFrom: .: {} f:resourceFieldRef: {} k:{"name":"ISTIO_META_APP_CONTAINERS"}: .: {} f:name: {} k:{"name":"ISTIO_META_CLUSTER_ID"}: .: {} f:name: {} f:value: {} k:{"name":"ISTIO_META_INTERCEPTION_MODE"}: .: {} f:name: {} f:value: {} k:{"name":"ISTIO_META_MESH_ID"}: .: {} f:name: {} f:value: {} k:{"name":"ISTIO_META_NODE_NAME"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"ISTIO_META_OWNER"}: .: {} f:name: {} f:value: {} k:{"name":"ISTIO_META_POD_PORTS"}: .: {} f:name: {} f:value: {} k:{"name":"ISTIO_META_WORKLOAD_NAME"}: .: {} f:name: {} f:value: {} k:{"name":"PILOT_CERT_PROVIDER"}: .: {} f:name: {} f:value: {} k:{"name":"POD_NAME"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"POD_NAMESPACE"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"PROXY_CONFIG"}: .: {} f:name: {} f:value: {} k:{"name":"SERVICE_ACCOUNT"}: .: {} f:name: {} f:valueFrom: .: {} f:fieldRef: {} k:{"name":"TRUST_DOMAIN"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:ports: .: {} k:{"containerPort":15020,"protocol":"TCP"}: .: {} f:containerPort: {} f:name: {} f:protocol: {} k:{"containerPort":15021,"protocol":"TCP"}: .: {} f:containerPort: {} f:name: {} f:protocol: {} k:{"containerPort":15090,"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:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:securityContext: .: {} f:allowPrivilegeEscalation: {} f:capabilities: .: {} f:drop: {} f:privileged: {} f:readOnlyRootFilesystem: {} f:runAsGroup: {} f:runAsNonRoot: {} f:runAsUser: {} 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":"/etc/istio/pod"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/etc/istio/proxy"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/lib/istio/data"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/secrets/credential-uds"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/secrets/istio"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/secrets/tokens"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/secrets/workload-spiffe-credentials"}: .: {} f:mountPath: {} f:name: {} k:{"mountPath":"/var/run/secrets/workload-spiffe-uds"}: .: {} f:mountPath: {} f:name: {} f:dnsPolicy: {} f:enableServiceLinks: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: .: {} f:sysctls: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"credential-socket"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"istio-data"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"istio-envoy"}: .: {} f:emptyDir: .: {} f:medium: {} f:name: {} k:{"name":"istio-podinfo"}: .: {} f:downwardAPI: .: {} f:defaultMode: {} f:items: {} f:name: {} k:{"name":"istio-token"}: .: {} f:name: {} f:projected: .: {} f:defaultMode: {} f:sources: {} k:{"name":"istiod-ca-cert"}: .: {} f:configMap: .: {} f:defaultMode: {} f:name: {} f:name: {} k:{"name":"workload-certs"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"workload-socket"}: .: {} f:emptyDir: {} f:name: {} manager: kube-controller-manager operation: Update time: "2026-07-09T18:44:51Z" - 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: "2026-07-09T18:44:51Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:status: f:conditions: k:{"type":"ContainersReady"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"Initialized"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodReadyToStartContainers"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} k:{"type":"PodScheduled"}: f:observedGeneration: {} k:{"type":"Ready"}: .: {} f:lastProbeTime: {} f:lastTransitionTime: {} f:observedGeneration: {} f:status: {} f:type: {} f:containerStatuses: {} f:hostIP: {} f:hostIPs: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.132.0.48"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-09T18:45:09Z" name: router-gateway-2-openshift-default-684dc9b4b9-r8qw5 namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: router-gateway-2-openshift-default-684dc9b4b9 uid: 5987bc37-5e76-466c-9d39-5ea3b765ad9e resourceVersion: "46083" uid: 44a89107-00ea-4275-ab07-7291b5b7e3c4 spec: containers: - args: - proxy - router - --domain - $(POD_NAMESPACE).svc.cluster.local - --proxyLogLevel - warning - --proxyComponentLogLevel - misc:error - --log_output_level - default:info env: - name: PILOT_CERT_PROVIDER value: istiod - name: CA_ADDR value: istiod-openshift-gateway.openshift-ingress.svc:15012 - name: POD_NAME valueFrom: fieldRef: apiVersion: v1 fieldPath: metadata.name - name: POD_NAMESPACE valueFrom: fieldRef: apiVersion: v1 fieldPath: metadata.namespace - name: INSTANCE_IP valueFrom: fieldRef: apiVersion: v1 fieldPath: status.podIP - name: SERVICE_ACCOUNT valueFrom: fieldRef: apiVersion: v1 fieldPath: spec.serviceAccountName - name: HOST_IP valueFrom: fieldRef: apiVersion: v1 fieldPath: status.hostIP - name: ISTIO_CPU_LIMIT valueFrom: resourceFieldRef: divisor: "0" resource: limits.cpu - name: PROXY_CONFIG value: | {"discoveryAddress":"istiod-openshift-gateway.openshift-ingress.svc:15012","proxyHeaders":{"server":{"disabled":true},"envoyDebugHeaders":{"disabled":true},"metadataExchangeHeaders":{"mode":"IN_MESH"}}} - name: ISTIO_META_POD_PORTS value: '[]' - name: ISTIO_META_APP_CONTAINERS - name: GOMEMLIMIT valueFrom: resourceFieldRef: divisor: "0" resource: limits.memory - name: GOMAXPROCS valueFrom: resourceFieldRef: divisor: "0" resource: limits.cpu - name: ISTIO_META_CLUSTER_ID value: Kubernetes - name: ISTIO_META_NODE_NAME valueFrom: fieldRef: apiVersion: v1 fieldPath: spec.nodeName - name: ISTIO_META_INTERCEPTION_MODE value: REDIRECT - name: ISTIO_META_WORKLOAD_NAME value: router-gateway-2-openshift-default - name: ISTIO_META_OWNER value: kubernetes://apis/apps/v1/namespaces/kserve-ci-e2e-test/deployments/router-gateway-2-openshift-default - name: ISTIO_META_MESH_ID value: cluster.local - name: TRUST_DOMAIN value: cluster.local image: registry.redhat.io/openshift-service-mesh/istio-proxyv2-rhel9@sha256:7d15cebf9b62f3f235c0eab5158ac8ff2fda86a1d193490dc94c301402c99da8 imagePullPolicy: IfNotPresent name: istio-proxy ports: - containerPort: 15020 name: metrics protocol: TCP - containerPort: 15021 name: status-port protocol: TCP - containerPort: 15090 name: http-envoy-prom protocol: TCP readinessProbe: failureThreshold: 4 httpGet: path: /healthz/ready port: 15021 scheme: HTTP periodSeconds: 15 successThreshold: 1 timeoutSeconds: 1 resources: limits: cpu: "2" memory: 2Gi requests: cpu: 100m memory: 256Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL privileged: false readOnlyRootFilesystem: true runAsGroup: 1000699999 runAsNonRoot: true runAsUser: 1000699999 startupProbe: failureThreshold: 30 httpGet: path: /healthz/ready port: 15021 scheme: HTTP initialDelaySeconds: 1 periodSeconds: 1 successThreshold: 1 timeoutSeconds: 1 terminationMessagePath: /dev/termination-log terminationMessagePolicy: File volumeMounts: - mountPath: /var/run/secrets/workload-spiffe-uds name: workload-socket - mountPath: /var/run/secrets/credential-uds name: credential-socket - mountPath: /var/run/secrets/workload-spiffe-credentials name: workload-certs - mountPath: /var/run/secrets/istio name: istiod-ca-cert - mountPath: /var/lib/istio/data name: istio-data - mountPath: /etc/istio/proxy name: istio-envoy - mountPath: /var/run/secrets/tokens name: istio-token - mountPath: /etc/istio/pod name: istio-podinfo - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-g972t readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: router-gateway-2-openshift-default-dockercfg-q8vnq nodeName: ip-10-0-137-79.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault sysctls: - name: net.ipv4.ip_unprivileged_port_start value: "0" serviceAccount: router-gateway-2-openshift-default serviceAccountName: router-gateway-2-openshift-default 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: - emptyDir: {} name: workload-socket - emptyDir: {} name: credential-socket - emptyDir: {} name: workload-certs - emptyDir: medium: Memory name: istio-envoy - emptyDir: {} name: istio-data - downwardAPI: defaultMode: 420 items: - fieldRef: apiVersion: v1 fieldPath: metadata.labels path: labels - fieldRef: apiVersion: v1 fieldPath: metadata.annotations path: annotations name: istio-podinfo - name: istio-token projected: defaultMode: 420 sources: - serviceAccountToken: audience: istio-ca expirationSeconds: 43200 path: istio-token - configMap: defaultMode: 420 name: openshift-gw-ca-root-cert name: istiod-ca-cert - name: kube-api-access-g972t 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: "2026-07-09T18:44:51Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-09T18:44:51Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-09T18:45:09Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-09T18:45:09Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-09T18:44:51Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 100m memory: 256Mi containerID: cri-o://9a65ff41377e0e331f7b94c228824a79dc8c553b50e474db45d016a4b958d56b image: registry.redhat.io/openshift-service-mesh/istio-proxyv2-rhel9@sha256:7d15cebf9b62f3f235c0eab5158ac8ff2fda86a1d193490dc94c301402c99da8 imageID: registry.redhat.io/openshift-service-mesh/istio-proxyv2-rhel9@sha256:0ae6667459cd9e9ea0549e27bb014c722924454bf594b31646df8df58be91cc5 lastState: terminated: containerID: cri-o://aae98ca7ea8951a481a0eebd7cafb06bd22925f4258b0d2ee24931efa1e3655d exitCode: 0 finishedAt: "2026-07-09T18:45:02Z" reason: Completed startedAt: "2026-07-09T18:44:51Z" name: istio-proxy ready: true resources: limits: cpu: "2" memory: 2Gi requests: cpu: 100m memory: 256Mi restartCount: 1 started: true state: running: startedAt: "2026-07-09T18:45:02Z" user: linux: gid: 1000699999 supplementalGroups: - 1000699999 - 1000690000 uid: 1000699999 volumeMounts: - mountPath: /var/run/secrets/workload-spiffe-uds name: workload-socket - mountPath: /var/run/secrets/credential-uds name: credential-socket - mountPath: /var/run/secrets/workload-spiffe-credentials name: workload-certs - mountPath: /var/run/secrets/istio name: istiod-ca-cert - mountPath: /var/lib/istio/data name: istio-data - mountPath: /etc/istio/proxy name: istio-envoy - mountPath: /var/run/secrets/tokens name: istio-token - mountPath: /etc/istio/pod name: istio-podinfo - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-g972t readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.137.79 hostIPs: - ip: 10.0.137.79 observedGeneration: 1 phase: Running podIP: 10.132.0.48 podIPs: - ip: 10.132.0.48 qosClass: Burstable startTime: "2026-07-09T18:44:51Z" kind: PodList metadata: resourceVersion: "81296"