--- apiVersion: v1 items: - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.134.0.32/23"],"mac_address":"0a:58:0a:86:00:20","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":"10.134.0.32/23","gateway_ip":"10.134.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.134.0.32" ], "mac": "0a:58:0a:86:00:20", "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: "2026-07-08T07:14:55Z" generateName: auth-disabled-test-kserve-68f5d8959c- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload app.kubernetes.io/name: auth-disabled-test app.kubernetes.io/part-of: llminferenceservice kserve.io/component: workload llm-d.ai/role: both pod-template-hash: 68f5d8959c managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-140-110 operation: Update subresource: status time: "2026-07-08T07:14:55Z" - 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":"594d98e8-b847-491f-bca7-d49afb026bb0"}: {} 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:initContainers: .: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"dshm"}: .: {} f:emptyDir: .: {} f:medium: {} f:sizeLimit: {} f:name: {} k:{"name":"home"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} 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: "2026-07-08T07:14:55Z" - 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: "2026-07-08T07:14:55Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.134.0.32"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:16:55Z" name: auth-disabled-test-kserve-68f5d8959c-xh45g namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: auth-disabled-test-kserve-68f5d8959c uid: 594d98e8-b847-491f-bca7-d49afb026bb0 resourceVersion: "30559" uid: 275c2425-5390-4987-804a-f70b52e9dd60 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-provision-location readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-8s824 readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw initContainers: - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-8s824 readOnly: true nodeName: ip-10-0-140-110.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: auth-disabled-test-kserve-self-signed-certs - emptyDir: {} name: kserve-provision-location - name: kube-api-access-8s824 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: "2026-07-08T07:14:56Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:15:01Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:16:55Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:16:55Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:14:55Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 200m memory: 2Gi containerID: cri-o://64b85f4df3d23c2ee6d0f99adcdcd2adac556b592bf194f03a5310372146755c image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo@sha256:afb39fca138b51d019d986229d546531b45a2a3deb73bcf59bd42406e13fbba0 lastState: {} name: main ready: true resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi restartCount: 0 started: true state: running: startedAt: "2026-07-08T07:15:01Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 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-provision-location readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-8s824 readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.140.110 hostIPs: - ip: 10.0.140.110 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://23bb9ea2aa028659882ccccb4e7044bd962e470f66de45ef1b29479af110f60b image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://23bb9ea2aa028659882ccccb4e7044bd962e470f66de45ef1b29479af110f60b exitCode: 0 finishedAt: "2026-07-08T07:15:00Z" reason: Completed startedAt: "2026-07-08T07:14:55Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-8s824 readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.134.0.32 podIPs: - ip: 10.134.0.32 qosClass: Burstable startTime: "2026-07-08T07:14:55Z" - 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":["10.134.0.33/23"],"mac_address":"0a:58:0a:86:00:21","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":"10.134.0.33/23","gateway_ip":"10.134.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.134.0.33" ], "mac": "0a:58:0a:86:00:21", "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: "2026-07-08T07:14:55Z" generateName: auth-disabled-test-kserve-router-scheduler-5898f785fb- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-router-scheduler app.kubernetes.io/name: auth-disabled-test app.kubernetes.io/part-of: llminferenceservice pod-template-hash: 5898f785fb managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-140-110 operation: Update subresource: status time: "2026-07-08T07:14:55Z" - 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":"0067b8f6-2d4d-4411-8c67-36525950f914"}: {} 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:initContainers: .: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} k:{"name":"STORAGE_ALLOW_PATTERNS"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tls-certs"}: .: {} f:name: {} f:secret: .: {} f:defaultMode: {} f:secretName: {} k:{"name":"tokenizer-cache"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-tmp"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-uds"}: .: {} f:emptyDir: {} f:name: {} manager: kube-controller-manager operation: Update time: "2026-07-08T07:14:55Z" - 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: "2026-07-08T07:14:56Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.134.0.33"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:15:30Z" name: auth-disabled-test-kserve-router-scheduler-5898f785fb-xr25g namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: auth-disabled-test-kserve-router-scheduler-5898f785fb uid: 0067b8f6-2d4d-4411-8c67-36525950f914 resourceVersion: "29131" uid: b0f58936-1d64-48ee-90e4-40b8b1145eeb 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 - auth-disabled-test-inference-pool - --pool-namespace - kserve-ci-e2e-test - --zap-encoder - json - --grpc-port - "9002" - --grpc-health-port - "9003" - --enable-cert-reload=true - --secure-serving=true - --model-server-metrics-scheme=https - --cert-path=/var/run/kserve/tls env: - name: SSL_CERT_DIR value: /var/run/kserve/tls:/var/run/secrets/kubernetes.io/serviceaccount:/etc/pki/tls/certs image: ghcr.io/llm-d/llm-d-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-sbcrz readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw - name: auth-disabled-test-epp-sa-dockercfg-r47px initContainers: - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" - name: STORAGE_ALLOW_PATTERNS value: '["tokenizer.json", "tokenizer_config.json", "special_tokens_map.json", "vocab.json", "merges.txt", "config.json", "generation_config.json"]' image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-sbcrz readOnly: true nodeName: ip-10-0-140-110.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: auth-disabled-test-epp-sa serviceAccountName: auth-disabled-test-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: auth-disabled-test-kserve-self-signed-certs - emptyDir: {} name: tokenizer-uds - emptyDir: {} name: tokenizer-tmp - emptyDir: {} name: tokenizer-cache - emptyDir: {} name: kserve-provision-location - name: kube-api-access-sbcrz 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: "2026-07-08T07:14:56Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:14:57Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:15:30Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:15:30Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:14:55Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 256m memory: 500Mi containerID: cri-o://f71b34a77ed9344f9c86cc6ba7ca22ec933a55f6a3ab320ce1b79866a42674d3 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: "2026-07-08T07:14:57Z" 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-sbcrz readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.140.110 hostIPs: - ip: 10.0.140.110 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://67fe7b435d75167fb6cad8724cee0589fa079c490b76eb61c465d3f48d2b5a33 image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://67fe7b435d75167fb6cad8724cee0589fa079c490b76eb61c465d3f48d2b5a33 exitCode: 0 finishedAt: "2026-07-08T07:14:56Z" reason: Completed startedAt: "2026-07-08T07:14:56Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-sbcrz readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.134.0.33 podIPs: - ip: 10.134.0.33 qosClass: Burstable startTime: "2026-07-08T07:14:55Z" - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.134.0.38/23"],"mac_address":"0a:58:0a:86:00:26","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":"10.134.0.38/23","gateway_ip":"10.134.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.134.0.38" ], "mac": "0a:58:0a:86:00:26", "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: "2026-07-08T07:38:21Z" generateName: custom-route-timeout-test-kserve-b6f9dd5f7- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload app.kubernetes.io/name: custom-route-timeout-test app.kubernetes.io/part-of: llminferenceservice kserve.io/component: workload llm-d.ai/role: both pod-template-hash: b6f9dd5f7 managedFields: - 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":"4ace3925-32b3-4a6b-851c-7e654e33fa07"}: {} 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:initContainers: .: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"dshm"}: .: {} f:emptyDir: .: {} f:medium: {} f:sizeLimit: {} f:name: {} k:{"name":"home"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} 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: "2026-07-08T07:38:21Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-140-110 operation: Update subresource: status time: "2026-07-08T07:38:22Z" - 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: "2026-07-08T07:38:22Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.134.0.38"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:40:32Z" name: custom-route-timeout-test-kserve-b6f9dd5f7-llmnq namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: custom-route-timeout-test-kserve-b6f9dd5f7 uid: 4ace3925-32b3-4a6b-851c-7e654e33fa07 resourceVersion: "53245" uid: 552f4e64-1b67-44ee-adf4-56fdad788beb 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-provision-location readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-ql2hm readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw initContainers: - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-ql2hm readOnly: true nodeName: ip-10-0-140-110.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: custom-route-timeout-test-kserve-self-signed-certs - emptyDir: {} name: kserve-provision-location - name: kube-api-access-ql2hm 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: "2026-07-08T07:38:22Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:38:26Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:40:32Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:40:32Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:38:21Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 200m memory: 2Gi containerID: cri-o://a04c5a372d0f050d217a4d7c20ac1caef2e8ff526ed1d5582686b1f0b81ed5fa image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo@sha256:afb39fca138b51d019d986229d546531b45a2a3deb73bcf59bd42406e13fbba0 lastState: {} name: main ready: true resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi restartCount: 0 started: true state: running: startedAt: "2026-07-08T07:38:27Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 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-provision-location readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-ql2hm readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.140.110 hostIPs: - ip: 10.0.140.110 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://b5713160351417943dd5e98f001ad4028a78b4e0315728bfb8f2eab91137feb3 image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://b5713160351417943dd5e98f001ad4028a78b4e0315728bfb8f2eab91137feb3 exitCode: 0 finishedAt: "2026-07-08T07:38:26Z" reason: Completed startedAt: "2026-07-08T07:38:22Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-ql2hm readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.134.0.38 podIPs: - ip: 10.134.0.38 qosClass: Burstable startTime: "2026-07-08T07:38:22Z" - 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":["10.132.0.45/23"],"mac_address":"0a:58:0a:84:00:2d","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":"10.132.0.45/23","gateway_ip":"10.132.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.132.0.45" ], "mac": "0a:58:0a:84:00:2d", "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: "2026-07-08T07:38:22Z" generateName: custom-route-timeout-test-kserve-router-scheduler-65985dc45b- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-router-scheduler app.kubernetes.io/name: custom-route-timeout-test app.kubernetes.io/part-of: llminferenceservice pod-template-hash: 65985dc45b managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-129-54 operation: Update subresource: status time: "2026-07-08T07:38:22Z" - 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":"856ae2cd-e8e4-4c39-bfd0-ef380c3d8d47"}: {} 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:initContainers: .: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} k:{"name":"STORAGE_ALLOW_PATTERNS"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tls-certs"}: .: {} f:name: {} f:secret: .: {} f:defaultMode: {} f:secretName: {} k:{"name":"tokenizer-cache"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-tmp"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-uds"}: .: {} f:emptyDir: {} f:name: {} manager: kube-controller-manager operation: Update time: "2026-07-08T07:38:22Z" - 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: "2026-07-08T07:38:22Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.132.0.45"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:38:55Z" name: custom-route-timeout-test-kserve-router-scheduler-65985dc4bddbc namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: custom-route-timeout-test-kserve-router-scheduler-65985dc45b uid: 856ae2cd-e8e4-4c39-bfd0-ef380c3d8d47 resourceVersion: "51951" uid: c9deabdb-8a2c-4081-9b23-e151d0275be9 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 - custom-route-timeout-test-inference-pool - --pool-namespace - kserve-ci-e2e-test - --zap-encoder - json - --grpc-port - "9002" - --grpc-health-port - "9003" - --enable-cert-reload=true - --secure-serving=true - --model-server-metrics-scheme=https - --cert-path=/var/run/kserve/tls env: - name: SSL_CERT_DIR value: /var/run/kserve/tls:/var/run/secrets/kubernetes.io/serviceaccount:/etc/pki/tls/certs image: ghcr.io/llm-d/llm-d-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-9652l readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw - name: custom-route-timeout-test-epp-sa-dockercfg-pqfb5 initContainers: - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" - name: STORAGE_ALLOW_PATTERNS value: '["tokenizer.json", "tokenizer_config.json", "special_tokens_map.json", "vocab.json", "merges.txt", "config.json", "generation_config.json"]' image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-9652l readOnly: true nodeName: ip-10-0-129-54.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: custom-route-timeout-test-epp-sa serviceAccountName: custom-route-timeout-test-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: custom-route-timeout-test-kserve-self-signed-certs - emptyDir: {} name: tokenizer-uds - emptyDir: {} name: tokenizer-tmp - emptyDir: {} name: tokenizer-cache - emptyDir: {} name: kserve-provision-location - name: kube-api-access-9652l 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: "2026-07-08T07:38:23Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:38:23Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:38:55Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:38:55Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:38:22Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 256m memory: 500Mi containerID: cri-o://54e4da05ac1f352a11802c23cecfe1cd956bd92421004e444cb17d187705dfb7 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: "2026-07-08T07:38:23Z" 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-9652l readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.129.54 hostIPs: - ip: 10.0.129.54 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://44c7909793090dc28d9c83088a302634c290c1c411335157104ed52ef04710e4 image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://44c7909793090dc28d9c83088a302634c290c1c411335157104ed52ef04710e4 exitCode: 0 finishedAt: "2026-07-08T07:38:23Z" reason: Completed startedAt: "2026-07-08T07:38:22Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-9652l readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.132.0.45 podIPs: - ip: 10.132.0.45 qosClass: Burstable startTime: "2026-07-08T07:38:22Z" - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.132.0.44/23"],"mac_address":"0a:58:0a:84:00:2c","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":"10.132.0.44/23","gateway_ip":"10.132.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.132.0.44" ], "mac": "0a:58:0a:84:00:2c", "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: "2026-07-08T07:36:39Z" generateName: e2e-pvc-model-download- generation: 1 labels: batch.kubernetes.io/controller-uid: ec6fd2a6-c694-48a6-afb7-dd87c7b3542f batch.kubernetes.io/job-name: e2e-pvc-model-download controller-uid: ec6fd2a6-c694-48a6-afb7-dd87c7b3542f 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":"ec6fd2a6-c694-48a6-afb7-dd87c7b3542f"}: {} 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: "2026-07-08T07:36:39Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-129-54 operation: Update subresource: status time: "2026-07-08T07:36:42Z" - 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: "2026-07-08T07:36:52Z" - 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":"10.132.0.44"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:37:00Z" name: e2e-pvc-model-download-vfrk4 namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: batch/v1 blockOwnerDeletion: true controller: true kind: Job name: e2e-pvc-model-download uid: ec6fd2a6-c694-48a6-afb7-dd87c7b3542f resourceVersion: "50039" uid: 5837851c-03f6-49c9-bce4-9ef14221d0a8 spec: containers: - args: - hf://facebook/opt-125m - /mnt/models image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 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-9wxhw readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw nodeName: ip-10-0-129-54.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-9wxhw 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: "2026-07-08T07:37:00Z" observedGeneration: 1 status: "False" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:36:42Z" observedGeneration: 1 reason: PodCompleted status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:36:59Z" observedGeneration: 1 reason: PodCompleted status: "False" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:36:59Z" observedGeneration: 1 reason: PodCompleted status: "False" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:36:42Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - containerID: cri-o://53036fe7a3f252c8e663360db2467ff31efeb513728476f7e24d894ec23513f9 image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: false resources: {} restartCount: 0 started: false state: terminated: containerID: cri-o://53036fe7a3f252c8e663360db2467ff31efeb513728476f7e24d894ec23513f9 exitCode: 0 finishedAt: "2026-07-08T07:36:58Z" reason: Completed startedAt: "2026-07-08T07:36:52Z" 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-9wxhw readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.129.54 hostIPs: - ip: 10.0.129.54 observedGeneration: 1 phase: Succeeded podIP: 10.132.0.44 podIPs: - ip: 10.132.0.44 qosClass: BestEffort startTime: "2026-07-08T07:36:42Z" - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.134.0.35/23"],"mac_address":"0a:58:0a:86:00:23","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":"10.134.0.35/23","gateway_ip":"10.134.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.134.0.35" ], "mac": "0a:58:0a:86:00:23", "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: "2026-07-08T07:21:01Z" generateName: llmisvc-model-fb-opt-125m-route-87882a8e-kserve-5c57b59c67- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload app.kubernetes.io/name: llmisvc-model-fb-opt-125m-route-87882a8e app.kubernetes.io/part-of: llminferenceservice kserve.io/component: workload llm-d.ai/role: both pod-template-hash: 5c57b59c67 managedFields: - 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":"3d7e956d-0178-41fc-bb8a-a23beb61cfad"}: {} 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:initContainers: .: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"dshm"}: .: {} f:emptyDir: .: {} f:medium: {} f:sizeLimit: {} f:name: {} k:{"name":"home"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} 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: "2026-07-08T07:21:01Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-140-110 operation: Update subresource: status time: "2026-07-08T07:21:02Z" - 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: "2026-07-08T07:21:02Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.134.0.35"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:23:02Z" name: llmisvc-model-fb-opt-125m-route-87882a8e-kserve-5c57b59c67dgldz namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: llmisvc-model-fb-opt-125m-route-87882a8e-kserve-5c57b59c67 uid: 3d7e956d-0178-41fc-bb8a-a23beb61cfad resourceVersion: "36764" uid: eda082cc-9a14-4867-8ac5-19d540f8ec41 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-provision-location readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-d8bcg readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw initContainers: - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-d8bcg readOnly: true nodeName: ip-10-0-140-110.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: llmisva690bbc929faec8bc98c767f16c003c1-kserve-self-signed-certs - emptyDir: {} name: kserve-provision-location - name: kube-api-access-d8bcg 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: "2026-07-08T07:21:03Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:21:07Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:23:02Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:23:02Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:21:02Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 200m memory: 2Gi containerID: cri-o://840eb40da17cb7b8270fed3f09c1595c07b1f3ee221ab8de93103fe49dd89223 image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo@sha256:afb39fca138b51d019d986229d546531b45a2a3deb73bcf59bd42406e13fbba0 lastState: {} name: main ready: true resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi restartCount: 0 started: true state: running: startedAt: "2026-07-08T07:21:07Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 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-provision-location readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-d8bcg readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.140.110 hostIPs: - ip: 10.0.140.110 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://02994340f26e2c4e7b58a3efa3a339b0b764be9d8e2233a3aec5e9030cf0e9c5 image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://02994340f26e2c4e7b58a3efa3a339b0b764be9d8e2233a3aec5e9030cf0e9c5 exitCode: 0 finishedAt: "2026-07-08T07:21:06Z" reason: Completed startedAt: "2026-07-08T07:21:02Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-d8bcg readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.134.0.35 podIPs: - ip: 10.134.0.35 qosClass: Burstable startTime: "2026-07-08T07:21:02Z" - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.134.0.34/23"],"mac_address":"0a:58:0a:86:00:22","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":"10.134.0.34/23","gateway_ip":"10.134.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.134.0.34" ], "mac": "0a:58:0a:86:00:22", "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: "2026-07-08T07:16:26Z" generateName: llmisvc-model-fb-opt-125m-with-7ca60146-kserve-6694784b54- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload app.kubernetes.io/name: llmisvc-model-fb-opt-125m-with-7ca60146 app.kubernetes.io/part-of: llminferenceservice kserve.io/component: workload llm-d.ai/role: both pod-template-hash: 6694784b54 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-140-110 operation: Update subresource: status time: "2026-07-08T07:16:26Z" - 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":"58082a3a-5863-45c7-b4e7-9da5e3412076"}: {} f:spec: f:containers: k:{"name":"main"}: .: {} f:args: {} 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"}: .: {} 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":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"CA_BUNDLE_CONFIGMAP_NAME"}: .: {} f:name: {} f:value: {} k:{"name":"CA_BUNDLE_VOLUME_MOUNT_POINT"}: .: {} f:name: {} f:value: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/etc/ssl/custom-certs"}: .: {} f:mountPath: {} f:name: {} f:readOnly: {} k:{"mountPath":"/mnt"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"cabundle-cert"}: .: {} f:configMap: .: {} f:defaultMode: {} f:name: {} f:name: {} k:{"name":"dshm"}: .: {} f:emptyDir: .: {} f:medium: {} f:sizeLimit: {} f:name: {} k:{"name":"home"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} 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: "2026-07-08T07:16:26Z" - 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: "2026-07-08T07:16:27Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.134.0.34"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:17:56Z" name: llmisvc-model-fb-opt-125m-with-7ca60146-kserve-6694784b54-grkws namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: llmisvc-model-fb-opt-125m-with-7ca60146-kserve-6694784b54 uid: 58082a3a-5863-45c7-b4e7-9da5e3412076 resourceVersion: "31656" uid: ec1d0b1a-7cd1-49d9-b280-796bd8f53eca spec: containers: - args: - --enable-lora - --lora-modules - '''{"name":"lora-adapter-1","path":"/mnt/lora/lora-adapter-1"}''' - '''{"name":"publishers/kserve-ci-e2e-test/models/lora-adapter-1","path":"/mnt/lora/lora-adapter-1"}''' 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 name: kserve-provision-location readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-vqrvk readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw initContainers: - args: - hf://facebook/opt-125m - /mnt/models - hf://edbeeching/opt-125m-lora - /mnt/lora/lora-adapter-1 env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" - name: CA_BUNDLE_CONFIGMAP_NAME value: odh-kserve-custom-ca-bundle - name: CA_BUNDLE_VOLUME_MOUNT_POINT value: /etc/ssl/custom-certs image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt name: kserve-provision-location - mountPath: /etc/ssl/custom-certs name: cabundle-cert readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-vqrvk readOnly: true nodeName: ip-10-0-140-110.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: llmisv3e414c2ba058a022dfd694dbcbac5b51-kserve-self-signed-certs - emptyDir: {} name: kserve-provision-location - configMap: defaultMode: 420 name: odh-kserve-custom-ca-bundle name: cabundle-cert - name: kube-api-access-vqrvk 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: "2026-07-08T07:16:27Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:16:31Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:17:56Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:17:56Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:16:26Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 200m memory: 2Gi containerID: cri-o://8389621bde1a4cda78271338d70f6ccf449aceaf5bb12758cc561b41ab3cee16 image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo@sha256:afb39fca138b51d019d986229d546531b45a2a3deb73bcf59bd42406e13fbba0 lastState: {} name: main ready: true resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi restartCount: 0 started: true state: running: startedAt: "2026-07-08T07:16:32Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 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 name: kserve-provision-location readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-vqrvk readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.140.110 hostIPs: - ip: 10.0.140.110 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://a4dab4dc745a8178aa2bbfb64122df941950699277a41f326a7c4af283354c57 image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://a4dab4dc745a8178aa2bbfb64122df941950699277a41f326a7c4af283354c57 exitCode: 0 finishedAt: "2026-07-08T07:16:31Z" reason: Completed startedAt: "2026-07-08T07:16:27Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt name: kserve-provision-location - mountPath: /etc/ssl/custom-certs name: cabundle-cert readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-vqrvk readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.134.0.34 podIPs: - ip: 10.134.0.34 qosClass: Burstable startTime: "2026-07-08T07:16:26Z" - 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":["10.132.0.42/23"],"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"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.132.0.42" ], "mac": "0a:58:0a:84:00:2a", "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: "2026-07-08T07:16:26Z" generateName: llmisvc-model-fb-opt-125m-with-7ca60146-kserve-router-scheduler-6588f4987- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-router-scheduler app.kubernetes.io/name: llmisvc-model-fb-opt-125m-with-7ca60146 app.kubernetes.io/part-of: llminferenceservice pod-template-hash: 6588f4987 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-129-54 operation: Update subresource: status time: "2026-07-08T07:16:26Z" - 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":"91372b18-811f-4ca0-9afc-bbd8326478d0"}: {} 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:initContainers: .: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} k:{"name":"STORAGE_ALLOW_PATTERNS"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tls-certs"}: .: {} f:name: {} f:secret: .: {} f:defaultMode: {} f:secretName: {} k:{"name":"tokenizer-cache"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-tmp"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-uds"}: .: {} f:emptyDir: {} f:name: {} manager: kube-controller-manager operation: Update time: "2026-07-08T07:16:26Z" - 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: "2026-07-08T07:16:27Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.132.0.42"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:17:00Z" name: llmisvc-model-fb-opt-125m-with-7ca60146-kserve-router-schenqx6z namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: llmisvc-model-fb-opt-125m-with-7ca60146-kserve-router-scheduler-6588f4987 uid: 91372b18-811f-4ca0-9afc-bbd8326478d0 resourceVersion: "30645" uid: 7b24e8e6-7f75-4a57-a998-e2c566c96672 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 - type: lora-affinity-scorer schedulingProfiles: - name: default plugins: - pluginRef: lora-affinity-scorer weight: 4 - pluginRef: queue-scorer weight: 2 - pluginRef: prefix-cache-scorer weight: 3 - pluginRef: max-score-picker command: - /app/epp - --pool-name - llmisvc-model-fb-opt-125m-with-7ca60146-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-bqttg readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw - name: llmisvc-model-fb-opt-125m-with-7ca60146-epp-sa-dockercfg-z28xr initContainers: - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" - name: STORAGE_ALLOW_PATTERNS value: '["tokenizer.json", "tokenizer_config.json", "special_tokens_map.json", "vocab.json", "merges.txt", "config.json", "generation_config.json"]' image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-bqttg readOnly: true nodeName: ip-10-0-129-54.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-fb-opt-125m-with-7ca60146-epp-sa serviceAccountName: llmisvc-model-fb-opt-125m-with-7ca60146-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: llmisv3e414c2ba058a022dfd694dbcbac5b51-kserve-self-signed-certs - emptyDir: {} name: tokenizer-uds - emptyDir: {} name: tokenizer-tmp - emptyDir: {} name: tokenizer-cache - emptyDir: {} name: kserve-provision-location - name: kube-api-access-bqttg 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: "2026-07-08T07:16:28Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:16:28Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:17:00Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:17:00Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:16:26Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 256m memory: 500Mi containerID: cri-o://483ea1d8ba85297c14764a994bba78acac3bae045901455c6c5eecee98b7ceac 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: "2026-07-08T07:16:28Z" 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-bqttg readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.129.54 hostIPs: - ip: 10.0.129.54 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://0b79466f079583e4489f430f394e5dd146d4e3928cb18385f31050fb847c0e84 image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://0b79466f079583e4489f430f394e5dd146d4e3928cb18385f31050fb847c0e84 exitCode: 0 finishedAt: "2026-07-08T07:16:28Z" reason: Completed startedAt: "2026-07-08T07:16:27Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-bqttg readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.132.0.42 podIPs: - ip: 10.132.0.42 qosClass: Burstable startTime: "2026-07-08T07:16:26Z" - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.134.0.36/23"],"mac_address":"0a:58:0a:86:00:24","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":"10.134.0.36/23","gateway_ip":"10.134.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.134.0.36" ], "mac": "0a:58:0a:86:00:24", "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: "2026-07-08T07:33:12Z" generateName: llmisvc-model-fb-opt-125m-with-ba4d693a-kserve-55585bf4d9- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload app.kubernetes.io/name: llmisvc-model-fb-opt-125m-with-ba4d693a app.kubernetes.io/part-of: llminferenceservice kserve.io/component: workload llm-d.ai/role: both pod-template-hash: 55585bf4d9 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-140-110 operation: Update subresource: status time: "2026-07-08T07:33:12Z" - 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":"820401f2-a724-4a90-9f00-21df066ddbb6"}: {} f:spec: f:containers: k:{"name":"main"}: .: {} f:args: {} 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"}: .: {} 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":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"CA_BUNDLE_CONFIGMAP_NAME"}: .: {} f:name: {} f:value: {} k:{"name":"CA_BUNDLE_VOLUME_MOUNT_POINT"}: .: {} f:name: {} f:value: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/etc/ssl/custom-certs"}: .: {} f:mountPath: {} f:name: {} f:readOnly: {} k:{"mountPath":"/mnt"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"cabundle-cert"}: .: {} f:configMap: .: {} f:defaultMode: {} f:name: {} f:name: {} k:{"name":"dshm"}: .: {} f:emptyDir: .: {} f:medium: {} f:sizeLimit: {} f:name: {} k:{"name":"home"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} 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: "2026-07-08T07:33:12Z" - 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: "2026-07-08T07:33:12Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.134.0.36"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:34:32Z" name: llmisvc-model-fb-opt-125m-with-ba4d693a-kserve-55585bf4d9-8bdd6 namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: llmisvc-model-fb-opt-125m-with-ba4d693a-kserve-55585bf4d9 uid: 820401f2-a724-4a90-9f00-21df066ddbb6 resourceVersion: "47877" uid: 57b1e783-1750-4f22-8ebd-bc305536de13 spec: containers: - args: - --enable-lora - --lora-modules - '''{"name":"lora-adapter-1","path":"/mnt/lora/lora-adapter-1"}''' - '''{"name":"publishers/kserve-ci-e2e-test/models/lora-adapter-1","path":"/mnt/lora/lora-adapter-1"}''' 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 name: kserve-provision-location readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-48v8h readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw initContainers: - args: - hf://facebook/opt-125m - /mnt/models - hf://edbeeching/opt-125m-lora - /mnt/lora/lora-adapter-1 env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" - name: CA_BUNDLE_CONFIGMAP_NAME value: odh-kserve-custom-ca-bundle - name: CA_BUNDLE_VOLUME_MOUNT_POINT value: /etc/ssl/custom-certs image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt name: kserve-provision-location - mountPath: /etc/ssl/custom-certs name: cabundle-cert readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-48v8h readOnly: true nodeName: ip-10-0-140-110.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: llmisv77ff2528d3e9b4972cd9335229fce9f0-kserve-self-signed-certs - emptyDir: {} name: kserve-provision-location - configMap: defaultMode: 420 name: odh-kserve-custom-ca-bundle name: cabundle-cert - name: kube-api-access-48v8h 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: "2026-07-08T07:33:13Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:33:17Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:34:32Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:34:32Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:33:12Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 200m memory: 2Gi containerID: cri-o://069334125f3f3bed0010fe6043f1b884e6ca3f3f34b913a48f9e8cffb8376206 image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo@sha256:afb39fca138b51d019d986229d546531b45a2a3deb73bcf59bd42406e13fbba0 lastState: {} name: main ready: true resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi restartCount: 0 started: true state: running: startedAt: "2026-07-08T07:33:17Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 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 name: kserve-provision-location readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-48v8h readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.140.110 hostIPs: - ip: 10.0.140.110 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://65a1fb2b67a4d58bc0471ddb72ac701aaecb0b2070d06a1afed2c83a6624de9f image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://65a1fb2b67a4d58bc0471ddb72ac701aaecb0b2070d06a1afed2c83a6624de9f exitCode: 0 finishedAt: "2026-07-08T07:33:16Z" reason: Completed startedAt: "2026-07-08T07:33:13Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt name: kserve-provision-location - mountPath: /etc/ssl/custom-certs name: cabundle-cert readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-48v8h readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.134.0.36 podIPs: - ip: 10.134.0.36 qosClass: Burstable startTime: "2026-07-08T07:33:12Z" - 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":["10.132.0.43/23"],"mac_address":"0a:58:0a:84:00:2b","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":"10.132.0.43/23","gateway_ip":"10.132.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.132.0.43" ], "mac": "0a:58:0a:84:00:2b", "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: "2026-07-08T07:33:12Z" generateName: llmisvc-model-fb-opt-125m-with-ba4d693a-kserve-router-scheduler-5d6c88c6f4- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-router-scheduler app.kubernetes.io/name: llmisvc-model-fb-opt-125m-with-ba4d693a app.kubernetes.io/part-of: llminferenceservice pod-template-hash: 5d6c88c6f4 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-129-54 operation: Update subresource: status time: "2026-07-08T07:33:12Z" - 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":"63184586-2b54-4322-b829-3b08a81be636"}: {} 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:initContainers: .: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} k:{"name":"STORAGE_ALLOW_PATTERNS"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tls-certs"}: .: {} f:name: {} f:secret: .: {} f:defaultMode: {} f:secretName: {} k:{"name":"tokenizer-cache"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-tmp"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-uds"}: .: {} f:emptyDir: {} f:name: {} manager: kube-controller-manager operation: Update time: "2026-07-08T07:33:12Z" - 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: "2026-07-08T07:33:13Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.132.0.43"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:33:47Z" name: llmisvc-model-fb-opt-125m-with-ba4d693a-kserve-router-sche92hd4 namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: llmisvc-model-fb-opt-125m-with-ba4d693a-kserve-router-scheduler-5d6c88c6f4 uid: 63184586-2b54-4322-b829-3b08a81be636 resourceVersion: "47218" uid: 685b6e6b-dedb-4dc0-ba22-2711b4b10f95 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 - type: lora-affinity-scorer schedulingProfiles: - name: default plugins: - pluginRef: lora-affinity-scorer weight: 4 - pluginRef: queue-scorer weight: 2 - pluginRef: prefix-cache-scorer weight: 3 - pluginRef: max-score-picker command: - /app/epp - --pool-name - llmisvc-model-fb-opt-125m-with-ba4d693a-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-xgx84 readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw - name: llmisvc-model-fb-opt-125m-with-ba4d693a-epp-sa-dockercfg-gp76f initContainers: - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" - name: STORAGE_ALLOW_PATTERNS value: '["tokenizer.json", "tokenizer_config.json", "special_tokens_map.json", "vocab.json", "merges.txt", "config.json", "generation_config.json"]' image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-xgx84 readOnly: true nodeName: ip-10-0-129-54.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-fb-opt-125m-with-ba4d693a-epp-sa serviceAccountName: llmisvc-model-fb-opt-125m-with-ba4d693a-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: llmisv77ff2528d3e9b4972cd9335229fce9f0-kserve-self-signed-certs - emptyDir: {} name: tokenizer-uds - emptyDir: {} name: tokenizer-tmp - emptyDir: {} name: tokenizer-cache - emptyDir: {} name: kserve-provision-location - name: kube-api-access-xgx84 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: "2026-07-08T07:33:14Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:33:15Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:33:47Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:33:47Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:33:12Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 256m memory: 500Mi containerID: cri-o://91a693fc1f7c846b143f5db47763719b0f2c6877c091707b900c7cadcc15f211 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: "2026-07-08T07:33:15Z" 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-xgx84 readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.129.54 hostIPs: - ip: 10.0.129.54 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://724c09503ef09af5db70189ab0c45a67124493884a5222fa8b8cc01d0a13448b image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://724c09503ef09af5db70189ab0c45a67124493884a5222fa8b8cc01d0a13448b exitCode: 0 finishedAt: "2026-07-08T07:33:14Z" reason: Completed startedAt: "2026-07-08T07:33:13Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-xgx84 readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.132.0.43 podIPs: - ip: 10.132.0.43 qosClass: Burstable startTime: "2026-07-08T07:33:12Z" - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.134.0.41/23"],"mac_address":"0a:58:0a:86:00:29","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":"10.134.0.41/23","gateway_ip":"10.134.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.134.0.41" ], "mac": "0a:58:0a:86:00:29", "default": true, "dns": {} }] leaderworkerset.sigs.k8s.io/size: "2" openshift.io/scc: openshift-ai-llminferenceservice-scc seccomp.security.alpha.kubernetes.io/pod: runtime/default security.openshift.io/validated-scc-subject-type: serviceaccount creationTimestamp: "2026-07-08T07:54:26Z" generateName: llmisvc-model-pvc-router-manage-2577e794-kserve-mn- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload-leader app.kubernetes.io/name: llmisvc-model-pvc-router-manage-2577e794 app.kubernetes.io/part-of: llminferenceservice apps.kubernetes.io/pod-index: "0" controller-revision-hash: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-7b97d8cc8b kserve.io/component: workload leaderworkerset.sigs.k8s.io/group-index: "0" leaderworkerset.sigs.k8s.io/group-key: dc52feed9d7e0b6abd87bf4d45c9409765943cd8 leaderworkerset.sigs.k8s.io/name: llmisvc-model-pvc-router-manage-2577e794-kserve-mn leaderworkerset.sigs.k8s.io/template-revision-hash: 686d468674 leaderworkerset.sigs.k8s.io/worker-index: "0" llm-d.ai/role: both statefulset.kubernetes.io/pod-name: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-0 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-140-110 operation: Update subresource: status time: "2026-07-08T07:54:26Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: .: {} f:leaderworkerset.sigs.k8s.io/size: {} f:generateName: {} f:labels: .: {} f:app.kubernetes.io/component: {} f:app.kubernetes.io/name: {} f:app.kubernetes.io/part-of: {} f:apps.kubernetes.io/pod-index: {} f:controller-revision-hash: {} f:kserve.io/component: {} f:leaderworkerset.sigs.k8s.io/name: {} f:leaderworkerset.sigs.k8s.io/template-revision-hash: {} f:leaderworkerset.sigs.k8s.io/worker-index: {} f:llm-d.ai/role: {} f:statefulset.kubernetes.io/pod-name: {} f:ownerReferences: .: {} k:{"uid":"a8b1a88a-3580-4ae0-99a7-48f84bc43323"}: {} f:spec: f:containers: k:{"name":"main"}: .: {} f:args: {} 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:add: {} 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:hostname: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:subdomain: {} 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: "2026-07-08T07:54:26Z" - 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: "2026-07-08T07:54:27Z" - 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":"10.134.0.41"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:57:07Z" name: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-0 namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: StatefulSet name: llmisvc-model-pvc-router-manage-2577e794-kserve-mn uid: a8b1a88a-3580-4ae0-99a7-48f84bc43323 resourceVersion: "66944" uid: 9d37b6b1-ed42-4a7e-aa20-371c401b8c99 spec: containers: - args: - --served-model-name - facebook/opt-125m - --port - "8000" - --enable-ssl-refresh - --ssl-certfile - /var/run/kserve/tls/tls.crt - --ssl-keyfile - /var/run/kserve/tls/tls.key command: - vllm - serve - /mnt/models env: - name: LWS_LEADER_ADDRESS value: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-0.llmisvc-model-pvc-router-manage-2577e794-kserve-mn.kserve-ci-e2e-test - name: LWS_GROUP_SIZE value: "2" - name: LWS_WORKER_INDEX value: "0" - 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: HOME value: /home - name: VLLM_LOGGING_LEVEL value: INFO - 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: add: - IPC_LOCK - SYS_RAWIO - NET_RAW 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-cgrz9 readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true hostname: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-0 imagePullSecrets: - name: default-dockercfg-psjvw - name: llmisvc-model-pvc-router-manage-2577e7-b09f84b1-dockercfg-zg5nt nodeName: ip-10-0-140-110.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: llmisvc-model-pvc-router-manage-2577e794-kserve-mn serviceAccountName: llmisvc-model-pvc-router-manage-2577e794-kserve-mn subdomain: llmisvc-model-pvc-router-manage-2577e794-kserve-mn 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: 8Gi name: dshm - emptyDir: {} name: model-cache - name: tls-certs secret: defaultMode: 420 secretName: llmisv1dc4269d1ada5f2d28562215d180c57f-kserve-self-signed-certs - name: kserve-pvc-source persistentVolumeClaim: claimName: e2e-pvc-model-storage - name: kube-api-access-cgrz9 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: "2026-07-08T07:54:27Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:54:26Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:57:07Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:57:07Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:54:26Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 200m memory: 2Gi containerID: cri-o://1351ed9f7de239c3fb33203448f37b552ad981cd820bd7d4feb0762b1d07a587 image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo@sha256:afb39fca138b51d019d986229d546531b45a2a3deb73bcf59bd42406e13fbba0 lastState: {} name: main ready: true resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi restartCount: 0 started: true state: running: startedAt: "2026-07-08T07:54:27Z" user: linux: gid: 0 supplementalGroups: - 0 uid: 1000690000 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-cgrz9 readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.140.110 hostIPs: - ip: 10.0.140.110 observedGeneration: 1 phase: Running podIP: 10.134.0.41 podIPs: - ip: 10.134.0.41 qosClass: Burstable startTime: "2026-07-08T07:54:26Z" - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.134.0.42/23"],"mac_address":"0a:58:0a:86:00:2a","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":"10.134.0.42/23","gateway_ip":"10.134.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.134.0.42" ], "mac": "0a:58:0a:86:00:2a", "default": true, "dns": {} }] leaderworkerset.sigs.k8s.io/leader-name: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-0 leaderworkerset.sigs.k8s.io/size: "2" openshift.io/scc: openshift-ai-llminferenceservice-scc seccomp.security.alpha.kubernetes.io/pod: runtime/default security.openshift.io/validated-scc-subject-type: serviceaccount creationTimestamp: "2026-07-08T07:54:26Z" generateName: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-0- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload-worker app.kubernetes.io/name: llmisvc-model-pvc-router-manage-2577e794 app.kubernetes.io/part-of: llminferenceservice apps.kubernetes.io/pod-index: "1" controller-revision-hash: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-0-579bcdf848 leaderworkerset.sigs.k8s.io/group-index: "0" leaderworkerset.sigs.k8s.io/group-key: dc52feed9d7e0b6abd87bf4d45c9409765943cd8 leaderworkerset.sigs.k8s.io/name: llmisvc-model-pvc-router-manage-2577e794-kserve-mn leaderworkerset.sigs.k8s.io/template-revision-hash: 686d468674 leaderworkerset.sigs.k8s.io/worker-index: "1" statefulset.kubernetes.io/pod-name: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-0-1 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-140-110 operation: Update subresource: status time: "2026-07-08T07:54:26Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: .: {} f:leaderworkerset.sigs.k8s.io/leader-name: {} f:leaderworkerset.sigs.k8s.io/size: {} f:generateName: {} f:labels: .: {} f:app.kubernetes.io/component: {} f:app.kubernetes.io/name: {} f:app.kubernetes.io/part-of: {} f:apps.kubernetes.io/pod-index: {} f:controller-revision-hash: {} f:leaderworkerset.sigs.k8s.io/group-index: {} f:leaderworkerset.sigs.k8s.io/group-key: {} f:leaderworkerset.sigs.k8s.io/name: {} f:leaderworkerset.sigs.k8s.io/template-revision-hash: {} f:statefulset.kubernetes.io/pod-name: {} f:ownerReferences: .: {} k:{"uid":"bf45adb8-e68e-49ed-8a62-a9f52a930d75"}: {} f:spec: f:containers: k:{"name":"main"}: .: {} f:args: {} 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:name: {} f:ports: .: {} k:{"containerPort":8000,"protocol":"TCP"}: .: {} f:containerPort: {} f:protocol: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:securityContext: .: {} f:allowPrivilegeEscalation: {} f:capabilities: .: {} f:add: {} f:drop: {} f:readOnlyRootFilesystem: {} f:runAsNonRoot: {} f:seccompProfile: .: {} f:type: {} 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:hostname: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:subdomain: {} 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: "2026-07-08T07:54:26Z" - 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":"10.134.0.42"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:54:27Z" - 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: "2026-07-08T07:54:27Z" name: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-0-1 namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: StatefulSet name: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-0 uid: bf45adb8-e68e-49ed-8a62-a9f52a930d75 resourceVersion: "64666" uid: 48d4f055-cfec-400b-ac7c-666b216ccaca spec: containers: - args: - --served-model-name - facebook/opt-125m - --port - "8000" - --enable-ssl-refresh - --ssl-certfile - /var/run/kserve/tls/tls.crt - --ssl-keyfile - /var/run/kserve/tls/tls.key command: - vllm - serve - /mnt/models env: - name: LWS_LEADER_ADDRESS value: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-0.llmisvc-model-pvc-router-manage-2577e794-kserve-mn.kserve-ci-e2e-test - name: LWS_GROUP_SIZE value: "2" - name: LWS_WORKER_INDEX value: "1" - 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: HOME value: /home - name: VLLM_LOGGING_LEVEL value: INFO - 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" name: main ports: - containerPort: 8000 protocol: TCP resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi securityContext: allowPrivilegeEscalation: false capabilities: add: - IPC_LOCK - SYS_RAWIO - NET_RAW drop: - ALL readOnlyRootFilesystem: true runAsNonRoot: true runAsUser: 1000690000 seccompProfile: type: RuntimeDefault 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-4kqnx readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true hostname: llmisvc-model-pvc-router-manage-2577e794-kserve-mn-0-1 imagePullSecrets: - name: default-dockercfg-psjvw - name: llmisvc-model-pvc-router-manage-2577e7-b09f84b1-dockercfg-zg5nt nodeName: ip-10-0-140-110.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: llmisvc-model-pvc-router-manage-2577e794-kserve-mn serviceAccountName: llmisvc-model-pvc-router-manage-2577e794-kserve-mn subdomain: llmisvc-model-pvc-router-manage-2577e794-kserve-mn 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: 8Gi name: dshm - emptyDir: {} name: model-cache - name: tls-certs secret: defaultMode: 420 secretName: llmisv1dc4269d1ada5f2d28562215d180c57f-kserve-self-signed-certs - name: kserve-pvc-source persistentVolumeClaim: claimName: e2e-pvc-model-storage - name: kube-api-access-4kqnx 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: "2026-07-08T07:54:27Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:54:26Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:54:27Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:54:27Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:54:26Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 200m memory: 2Gi containerID: cri-o://4fefe27e8657e159f787fcd96351a3114d85a7758fe093ca33084d2e7658317d image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo@sha256:afb39fca138b51d019d986229d546531b45a2a3deb73bcf59bd42406e13fbba0 lastState: {} name: main ready: true resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi restartCount: 0 started: true state: running: startedAt: "2026-07-08T07:54:27Z" user: linux: gid: 0 supplementalGroups: - 0 uid: 1000690000 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-4kqnx readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.140.110 hostIPs: - ip: 10.0.140.110 observedGeneration: 1 phase: Running podIP: 10.134.0.42 podIPs: - ip: 10.134.0.42 qosClass: Burstable startTime: "2026-07-08T07:54:26Z" - 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: "2026-07-08T07:54:26Z" generateName: llmisvc1dc4269d1ada5f2d28562215d180c57f-kserve-router-scheduler-7dbcb75dbc- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-router-scheduler app.kubernetes.io/name: llmisvc-model-pvc-router-manage-2577e794 app.kubernetes.io/part-of: llminferenceservice pod-template-hash: 7dbcb75dbc managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-136-129 operation: Update subresource: status time: "2026-07-08T07:54:26Z" - 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":"eb1a45f1-18a2-457e-8636-190d32b62411"}: {} 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: "2026-07-08T07:54:26Z" - 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: "2026-07-08T07:54:27Z" - 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: "2026-07-08T07:54:58Z" name: llmisvc1dc4269d1ada5f2d28562215d180c57f-kserve-router-schem2lrz namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: llmisvc1dc4269d1ada5f2d28562215d180c57f-kserve-router-scheduler-7dbcb75dbc uid: eb1a45f1-18a2-457e-8636-190d32b62411 resourceVersion: "65128" uid: e303ae4c-c54e-4b1f-b3d5-ef3a1457c985 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-2577e794-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-rdksh readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw - name: llmisvc-model-pvc-router-manage-2577e794-epp-sa-dockercfg-vc25d nodeName: ip-10-0-136-129.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-2577e794-epp-sa serviceAccountName: llmisvc-model-pvc-router-manage-2577e794-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: llmisv1dc4269d1ada5f2d28562215d180c57f-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-rdksh 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: "2026-07-08T07:54:27Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:54:26Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:54:58Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:54:58Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:54:26Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 256m memory: 500Mi containerID: cri-o://77ddc2a3e8071bf770667487649bcad7336a608e563154afdc96d143743a2cd5 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: "2026-07-08T07:54:27Z" 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-rdksh readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.136.129 hostIPs: - ip: 10.0.136.129 observedGeneration: 1 phase: Running podIP: 10.133.0.51 podIPs: - ip: 10.133.0.51 qosClass: Burstable startTime: "2026-07-08T07:54:26Z" - 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":["10.133.0.48/23"],"mac_address":"0a:58:0a:85:00:30","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":"10.133.0.48/23","gateway_ip":"10.133.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.133.0.48" ], "mac": "0a:58:0a:85:00:30", "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: "2026-07-08T07:21:02Z" generateName: llmisvca690bbc929faec8bc98c767f16c003c1-kserve-router-scheduler-599658ff65- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-router-scheduler app.kubernetes.io/name: llmisvc-model-fb-opt-125m-route-87882a8e app.kubernetes.io/part-of: llminferenceservice pod-template-hash: 599658ff65 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-136-129 operation: Update subresource: status time: "2026-07-08T07:21:02Z" - 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":"112023ff-5d0f-4641-9e6f-569aa6a74f4c"}: {} 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:initContainers: .: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} k:{"name":"STORAGE_ALLOW_PATTERNS"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tls-certs"}: .: {} f:name: {} f:secret: .: {} f:defaultMode: {} f:secretName: {} k:{"name":"tokenizer-cache"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-tmp"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-uds"}: .: {} f:emptyDir: {} f:name: {} manager: kube-controller-manager operation: Update time: "2026-07-08T07:21:02Z" - 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: "2026-07-08T07:21:02Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.133.0.48"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:21:35Z" name: llmisvca690bbc929faec8bc98c767f16c003c1-kserve-router-sched2rsl namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: llmisvca690bbc929faec8bc98c767f16c003c1-kserve-router-scheduler-599658ff65 uid: 112023ff-5d0f-4641-9e6f-569aa6a74f4c resourceVersion: "35398" uid: f5902ce4-f2d8-4f73-9f5f-7e3576c09c21 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-fb-opt-125m-route-87882a8e-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-52b6r readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw - name: llmisvc-model-fb-opt-125m-route-87882a8e-epp-sa-dockercfg-skn5r initContainers: - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" - name: STORAGE_ALLOW_PATTERNS value: '["tokenizer.json", "tokenizer_config.json", "special_tokens_map.json", "vocab.json", "merges.txt", "config.json", "generation_config.json"]' image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-52b6r readOnly: true nodeName: ip-10-0-136-129.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-fb-opt-125m-route-87882a8e-epp-sa serviceAccountName: llmisvc-model-fb-opt-125m-route-87882a8e-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: llmisva690bbc929faec8bc98c767f16c003c1-kserve-self-signed-certs - emptyDir: {} name: tokenizer-uds - emptyDir: {} name: tokenizer-tmp - emptyDir: {} name: tokenizer-cache - emptyDir: {} name: kserve-provision-location - name: kube-api-access-52b6r 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: "2026-07-08T07:21:02Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:21:03Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:21:35Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:21:35Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:21:02Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 256m memory: 500Mi containerID: cri-o://4aed5a5329fea04204c3a969f9b9ebc0639a48ccedbbc739da433231b32d7482 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: "2026-07-08T07:21:03Z" 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-52b6r readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.136.129 hostIPs: - ip: 10.0.136.129 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://204b7baf0ba60749de99c2fbfb103ecb750fae5269b25b69dc8ad4e468c74581 image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://204b7baf0ba60749de99c2fbfb103ecb750fae5269b25b69dc8ad4e468c74581 exitCode: 0 finishedAt: "2026-07-08T07:21:03Z" reason: Completed startedAt: "2026-07-08T07:21:02Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-52b6r readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.133.0.48 podIPs: - ip: 10.133.0.48 qosClass: Burstable startTime: "2026-07-08T07:21:02Z" - apiVersion: v1 kind: Pod metadata: annotations: istio.io/rev: openshift-gateway k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.132.0.34/23"],"mac_address":"0a:58:0a:84:00:22","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":"10.132.0.34/23","gateway_ip":"10.132.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.132.0.34" ], "mac": "0a:58:0a:84:00:22", "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: "2026-07-08T07:07:58Z" generateName: router-gateway-1-openshift-default-799f46c59b- generation: 1 labels: gateway.istio.io/managed: istio.io-gateway-controller gateway.networking.k8s.io/gateway-name: router-gateway-1 pod-template-hash: 799f46c59b 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-10-0-129-54 operation: Update subresource: status time: "2026-07-08T07:07:58Z" - 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":"a6431df8-249b-4486-a74a-850f8322c05c"}: {} 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: "2026-07-08T07:07:58Z" - 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: "2026-07-08T07:07:58Z" - 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":"10.132.0.34"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:19:00Z" name: router-gateway-1-openshift-default-799f46c59b-kwjm2 namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: router-gateway-1-openshift-default-799f46c59b uid: a6431df8-249b-4486-a74a-850f8322c05c resourceVersion: "32809" uid: f353d789-6bd9-4653-a007-ac3dab8cbbe6 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: 1Gi requests: cpu: 100m memory: 128Mi 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-bsf25 readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: router-gateway-1-openshift-default-dockercfg-ctvf2 nodeName: ip-10-0-129-54.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-bsf25 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: "2026-07-08T07:08:01Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:07:58Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:19:00Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:19:00Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:07:58Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 100m memory: 128Mi containerID: cri-o://88576e42cc05072fc0babc912dbca6a6d74bcbd77df996378b5ecda4c10a9d00 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://a17ddf11ac1e6cdea0df6bde69da0aefb4ece25ed5cf27f574352bb5b5e1cbf4 exitCode: 0 finishedAt: "2026-07-08T07:18:28Z" reason: Completed startedAt: "2026-07-08T07:08:52Z" name: istio-proxy ready: true resources: limits: cpu: "2" memory: 1Gi requests: cpu: 100m memory: 128Mi restartCount: 3 started: true state: running: startedAt: "2026-07-08T07:18:53Z" 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-bsf25 readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.129.54 hostIPs: - ip: 10.0.129.54 observedGeneration: 1 phase: Running podIP: 10.132.0.34 podIPs: - ip: 10.132.0.34 qosClass: Burstable startTime: "2026-07-08T07:07:58Z" - apiVersion: v1 kind: Pod metadata: annotations: istio.io/rev: openshift-gateway k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.132.0.50/23"],"mac_address":"0a:58:0a:84:00:32","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":"10.132.0.50/23","gateway_ip":"10.132.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.132.0.50" ], "mac": "0a:58:0a:84:00:32", "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: "2026-07-08T08:18:28Z" generateName: router-gateway-2-openshift-default-54c789bdc6- generation: 1 labels: gateway.istio.io/managed: istio.io-gateway-controller gateway.networking.k8s.io/gateway-name: router-gateway-2 pod-template-hash: 54c789bdc6 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: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":"4a74aae6-460b-4bc4-bc4f-0a20c49d9aa8"}: {} 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: "2026-07-08T08:18:27Z" - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-129-54 operation: Update subresource: status time: "2026-07-08T08:18:28Z" - 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: "2026-07-08T08:18:28Z" - 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":"10.132.0.50"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T08:18:50Z" name: router-gateway-2-openshift-default-54c789bdc6-4nrs7 namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: router-gateway-2-openshift-default-54c789bdc6 uid: 4a74aae6-460b-4bc4-bc4f-0a20c49d9aa8 resourceVersion: "85052" uid: a3f9c188-9368-4945-8007-0d0d53a746ca 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: 1Gi requests: cpu: 100m memory: 128Mi 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-gvvdh readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: router-gateway-2-openshift-default-dockercfg-zfwm8 nodeName: ip-10-0-129-54.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-gvvdh 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: "2026-07-08T08:18:29Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T08:18:28Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T08:18:50Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T08:18:50Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T08:18:28Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 100m memory: 128Mi containerID: cri-o://ed722f7ac0b80046146f7d8c2a2540fa828c5c1d74ef87ac3755ae96d566809d 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://258c3f3419e58f72a472547464ab7fcdd48555f2993ad11ee2d2ccc5e28790dc exitCode: 0 finishedAt: "2026-07-08T08:18:43Z" reason: Completed startedAt: "2026-07-08T08:18:28Z" name: istio-proxy ready: true resources: limits: cpu: "2" memory: 1Gi requests: cpu: 100m memory: 128Mi restartCount: 1 started: true state: running: startedAt: "2026-07-08T08:18:43Z" 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-gvvdh readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.129.54 hostIPs: - ip: 10.0.129.54 observedGeneration: 1 phase: Running podIP: 10.132.0.50 podIPs: - ip: 10.132.0.50 qosClass: Burstable startTime: "2026-07-08T08:18:28Z" - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.133.0.58/23"],"mac_address":"0a:58:0a:85:00:3a","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":"10.133.0.58/23","gateway_ip":"10.133.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.133.0.58" ], "mac": "0a:58:0a:85:00:3a", "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: "2026-07-08T08:19:02Z" generateName: router-with-refs-pd-test-kserve-6978d748b8- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload app.kubernetes.io/name: router-with-refs-pd-test app.kubernetes.io/part-of: llminferenceservice kserve.io/component: workload llm-d.ai/role: decode pod-template-hash: 6978d748b8 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-136-129 operation: Update subresource: status time: "2026-07-08T08:19:02Z" - 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":"66fb1529-51cc-4b1e-b155-875b7876558b"}: {} 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: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"dshm"}: .: {} f:emptyDir: .: {} f:medium: {} f:sizeLimit: {} f:name: {} k:{"name":"home"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} 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: "2026-07-08T08:19:02Z" - 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: "2026-07-08T08:19:02Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.133.0.58"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T08:22:22Z" name: router-with-refs-pd-test-kserve-6978d748b8-9ssnf namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: router-with-refs-pd-test-kserve-6978d748b8 uid: 66fb1529-51cc-4b1e-b155-875b7876558b resourceVersion: "87977" uid: 6a8bb147-65b6-4266-9e9e-e69ffc0d2c74 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-provision-location readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-5x6ph readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw - name: router-with-refs-pd-test-kserve-dockercfg-vnl4j 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/router-with-refs-pd-test-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: router-with-refs-pd-test-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-5x6ph readOnly: true - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-5x6ph readOnly: true nodeName: ip-10-0-136-129.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: router-with-refs-pd-test-kserve serviceAccountName: router-with-refs-pd-test-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: router-with-refs-pd-test-kserve-self-signed-certs - emptyDir: {} name: kserve-provision-location - name: kube-api-access-5x6ph 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: "2026-07-08T08:19:04Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T08:19:08Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T08:22:22Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T08:22:22Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T08:19:02Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 200m memory: 2Gi containerID: cri-o://ae20dccd51b3425d9b14010c020209cd04fb075f4429fa3a6549c130f6766f23 image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo@sha256:afb39fca138b51d019d986229d546531b45a2a3deb73bcf59bd42406e13fbba0 lastState: {} name: main ready: true resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi restartCount: 0 started: true state: running: startedAt: "2026-07-08T08:19:08Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 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-provision-location readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-5x6ph readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.136.129 hostIPs: - ip: 10.0.136.129 initContainerStatuses: - containerID: cri-o://81ab1a7b1b02a7ff4762d704accc044c6f81402d8238b9ddb91e712b7bffab69 image: ghcr.io/llm-d/llm-d-router-disagg-sidecar:v0.9.0-rc.2 imageID: ghcr.io/llm-d/llm-d-router-disagg-sidecar@sha256:1ec2c035216ea0a4f8a6f7ccda55398fc3d9f71f487c135f49ba0dea08163f12 lastState: {} name: llm-d-routing-sidecar ready: true resources: {} restartCount: 0 started: true state: running: startedAt: "2026-07-08T08:19:03Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /var/run/kserve/tls name: tls-certs readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-5x6ph readOnly: true recursiveReadOnly: Disabled - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://001fed2ab22b916716fe3a627dafffc9d95bdb6ffbaee6e88fe8a7e8a3b37316 image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://001fed2ab22b916716fe3a627dafffc9d95bdb6ffbaee6e88fe8a7e8a3b37316 exitCode: 0 finishedAt: "2026-07-08T08:19:07Z" reason: Completed startedAt: "2026-07-08T08:19:04Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-5x6ph readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.133.0.58 podIPs: - ip: 10.133.0.58 qosClass: Burstable startTime: "2026-07-08T08:19:02Z" - apiVersion: v1 kind: Pod metadata: annotations: k8s.ovn.org/pod-networks: '{"default":{"ip_addresses":["10.133.0.59/23"],"mac_address":"0a:58:0a:85:00:3b","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":"10.133.0.59/23","gateway_ip":"10.133.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.133.0.59" ], "mac": "0a:58:0a:85:00:3b", "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: "2026-07-08T08:19:02Z" generateName: router-with-refs-pd-test-kserve-prefill-56559b5fdf- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload-prefill app.kubernetes.io/name: router-with-refs-pd-test app.kubernetes.io/part-of: llminferenceservice kserve.io/component: workload llm-d.ai/role: prefill pod-template-hash: 56559b5fdf managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-136-129 operation: Update subresource: status time: "2026-07-08T08:19:02Z" - 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":"4ea7d5dc-71b8-4bfd-8bc6-9c429ce26728"}: {} 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:initContainers: .: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"dshm"}: .: {} f:emptyDir: .: {} f:medium: {} f:sizeLimit: {} f:name: {} k:{"name":"home"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} 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: "2026-07-08T08:19:02Z" - 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: "2026-07-08T08:19:03Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.133.0.59"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T08:22:22Z" name: router-with-refs-pd-test-kserve-prefill-56559b5fdf-jgcng namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: router-with-refs-pd-test-kserve-prefill-56559b5fdf uid: 4ea7d5dc-71b8-4bfd-8bc6-9c429ce26728 resourceVersion: "87983" uid: 9c57983a-88f5-4615-9e6b-8bafc4a48dfa 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-provision-location readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-82ptc readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw initContainers: - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-82ptc readOnly: true nodeName: ip-10-0-136-129.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: router-with-refs-pd-test-kserve-self-signed-certs - emptyDir: {} name: kserve-provision-location - name: kube-api-access-82ptc 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: "2026-07-08T08:19:04Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T08:19:07Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T08:22:22Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T08:22:22Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T08:19:02Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 200m memory: 2Gi containerID: cri-o://2931b47c3713ae15299afbf496c6bf0d53c43a3284aac9a9d85cb5cda4f7c376 image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo@sha256:afb39fca138b51d019d986229d546531b45a2a3deb73bcf59bd42406e13fbba0 lastState: {} name: main ready: true resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi restartCount: 0 started: true state: running: startedAt: "2026-07-08T08:19:07Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 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-provision-location readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-82ptc readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.136.129 hostIPs: - ip: 10.0.136.129 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://651c0c4dce5c3fe6951ae3b19c10556363eeba9c8ce6cc8ca176cec5abbe4c02 image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://651c0c4dce5c3fe6951ae3b19c10556363eeba9c8ce6cc8ca176cec5abbe4c02 exitCode: 0 finishedAt: "2026-07-08T08:19:06Z" reason: Completed startedAt: "2026-07-08T08:19:03Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-82ptc readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.133.0.59 podIPs: - ip: 10.133.0.59 qosClass: Burstable startTime: "2026-07-08T08:19:02Z" - 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":["10.132.0.51/23"],"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"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.132.0.51" ], "mac": "0a:58:0a:84:00:33", "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: "2026-07-08T08:19:03Z" generateName: router-with-refs-pd-test-kserve-router-scheduler-6ddb8947b5- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-router-scheduler app.kubernetes.io/name: router-with-refs-pd-test app.kubernetes.io/part-of: llminferenceservice pod-template-hash: 6ddb8947b5 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-129-54 operation: Update subresource: status time: "2026-07-08T08:19:03Z" - 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":"b2a90872-a2e7-4f27-968f-5f43b2f0851c"}: {} 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:initContainers: .: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} k:{"name":"STORAGE_ALLOW_PATTERNS"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tls-certs"}: .: {} f:name: {} f:secret: .: {} f:defaultMode: {} f:secretName: {} k:{"name":"tokenizer-cache"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-tmp"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-uds"}: .: {} f:emptyDir: {} f:name: {} manager: kube-controller-manager operation: Update time: "2026-07-08T08:19:03Z" - 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: "2026-07-08T08:19:03Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.132.0.51"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T08:19:37Z" name: router-with-refs-pd-test-kserve-router-scheduler-6ddb8947brh5kh namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: router-with-refs-pd-test-kserve-router-scheduler-6ddb8947b5 uid: b2a90872-a2e7-4f27-968f-5f43b2f0851c resourceVersion: "85895" uid: 63405bd3-69a5-4e13-bdfd-f5f34ab6f635 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 - router-with-refs-pd-test-inference-pool - --pool-namespace - kserve-ci-e2e-test - --zap-encoder - json - --grpc-port - "9002" - --grpc-health-port - "9003" - --enable-cert-reload=true - --secure-serving=true - --model-server-metrics-scheme=https - --cert-path=/var/run/kserve/tls env: - name: SSL_CERT_DIR value: /var/run/kserve/tls:/var/run/secrets/kubernetes.io/serviceaccount:/etc/pki/tls/certs image: ghcr.io/llm-d/llm-d-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-lzhz2 readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw - name: router-with-refs-pd-test-epp-sa-dockercfg-9tlf6 initContainers: - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" - name: STORAGE_ALLOW_PATTERNS value: '["tokenizer.json", "tokenizer_config.json", "special_tokens_map.json", "vocab.json", "merges.txt", "config.json", "generation_config.json"]' image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-lzhz2 readOnly: true nodeName: ip-10-0-129-54.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: router-with-refs-pd-test-epp-sa serviceAccountName: router-with-refs-pd-test-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: router-with-refs-pd-test-kserve-self-signed-certs - emptyDir: {} name: tokenizer-uds - emptyDir: {} name: tokenizer-tmp - emptyDir: {} name: tokenizer-cache - emptyDir: {} name: kserve-provision-location - name: kube-api-access-lzhz2 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: "2026-07-08T08:19:04Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T08:19:05Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T08:19:37Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T08:19:37Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T08:19:03Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 256m memory: 500Mi containerID: cri-o://dbb7295d2e8563de5b7aabb0f2b63ad4a274856db4dc7d111cbbe5b002b26841 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: "2026-07-08T08:19:05Z" 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-lzhz2 readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.129.54 hostIPs: - ip: 10.0.129.54 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://fb4ec50659ff64095b0c78c7a47163703c70ff163441c31d27a4e1943210b009 image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://fb4ec50659ff64095b0c78c7a47163703c70ff163441c31d27a4e1943210b009 exitCode: 0 finishedAt: "2026-07-08T08:19:04Z" reason: Completed startedAt: "2026-07-08T08:19:03Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-lzhz2 readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.132.0.51 podIPs: - ip: 10.132.0.51 qosClass: Burstable startTime: "2026-07-08T08:19:03Z" - 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: "2026-07-08T07:55:58Z" generateName: router-with-refs-test-kserve-794b6f7745- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-workload app.kubernetes.io/name: router-with-refs-test app.kubernetes.io/part-of: llminferenceservice kserve.io/component: workload llm-d.ai/role: both pod-template-hash: 794b6f7745 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-140-110 operation: Update subresource: status time: "2026-07-08T07:55:58Z" - 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":"071e58f0-2946-4c4e-a449-a998abd15a39"}: {} 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:initContainers: .: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"dshm"}: .: {} f:emptyDir: .: {} f:medium: {} f:sizeLimit: {} f:name: {} k:{"name":"home"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} 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: "2026-07-08T07:55:58Z" - 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: "2026-07-08T07:55:59Z" - 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:initContainerStatuses: {} 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: "2026-07-08T07:58:09Z" name: router-with-refs-test-kserve-794b6f7745-fpslh namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: router-with-refs-test-kserve-794b6f7745 uid: 071e58f0-2946-4c4e-a449-a998abd15a39 resourceVersion: "67751" uid: df3d2e0b-ce6b-4275-b4f5-dae9bf0960e9 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-provision-location readOnly: true - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-49gkq readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw initContainers: - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-49gkq readOnly: true nodeName: ip-10-0-140-110.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: router-with-refs-test-kserve-self-signed-certs - emptyDir: {} name: kserve-provision-location - name: kube-api-access-49gkq 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: "2026-07-08T07:55:59Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:56:03Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:58:08Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:58:08Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:55:58Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 200m memory: 2Gi containerID: cri-o://c47414320f5d5c56f7263389be5bcf9e8d407d0d3b9f0551649b450fd4fd98ac image: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo:v0.19.0 imageID: public.ecr.aws/q9t5s3a7/vllm-cpu-release-repo@sha256:afb39fca138b51d019d986229d546531b45a2a3deb73bcf59bd42406e13fbba0 lastState: {} name: main ready: true resources: limits: cpu: "2" memory: 7Gi requests: cpu: 200m memory: 2Gi restartCount: 0 started: true state: running: startedAt: "2026-07-08T07:56:03Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 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-provision-location readOnly: true recursiveReadOnly: Disabled - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-49gkq readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.140.110 hostIPs: - ip: 10.0.140.110 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://259e99435fc59eaf20dcf4f5c9f9167468b55ca52f5247c7371ebe618081f111 image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://259e99435fc59eaf20dcf4f5c9f9167468b55ca52f5247c7371ebe618081f111 exitCode: 0 finishedAt: "2026-07-08T07:56:02Z" reason: Completed startedAt: "2026-07-08T07:55:59Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-49gkq readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.134.0.43 podIPs: - ip: 10.134.0.43 qosClass: Burstable startTime: "2026-07-08T07:55:58Z" - 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":["10.133.0.52/23"],"mac_address":"0a:58:0a:85:00:34","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":"10.133.0.52/23","gateway_ip":"10.133.0.1","role":"primary"}}' k8s.v1.cni.cncf.io/network-status: |- [{ "name": "ovn-kubernetes", "interface": "eth0", "ips": [ "10.133.0.52" ], "mac": "0a:58:0a:85:00:34", "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: "2026-07-08T07:55:59Z" generateName: router-with-refs-test-kserve-router-scheduler-74dd96b744- generation: 1 labels: app.kubernetes.io/component: llminferenceservice-router-scheduler app.kubernetes.io/name: router-with-refs-test app.kubernetes.io/part-of: llminferenceservice pod-template-hash: 74dd96b744 managedFields: - apiVersion: v1 fieldsType: FieldsV1 fieldsV1: f:metadata: f:annotations: f:k8s.ovn.org/pod-networks: {} manager: ip-10-0-136-129 operation: Update subresource: status time: "2026-07-08T07:55:59Z" - 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":"0bbf19de-e126-4d6c-8580-b83de6f3a968"}: {} 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:initContainers: .: {} k:{"name":"storage-initializer"}: .: {} f:args: {} f:env: .: {} k:{"name":"AWS_ACCESS_KEY_ID"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"AWS_CA_BUNDLE"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_CA_BUNDLE_CONFIGMAP"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_ENDPOINT_URL"}: .: {} f:name: {} f:value: {} k:{"name":"AWS_SECRET_ACCESS_KEY"}: .: {} f:name: {} f:valueFrom: .: {} f:secretKeyRef: {} k:{"name":"HF_HUB_ENABLE_HF_TRANSFER"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_HIGH_PERFORMANCE"}: .: {} f:name: {} f:value: {} k:{"name":"HF_XET_NUM_CONCURRENT_RANGE_GETS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_ENDPOINT"}: .: {} f:name: {} f:value: {} k:{"name":"S3_USE_HTTPS"}: .: {} f:name: {} f:value: {} k:{"name":"S3_VERIFY_SSL"}: .: {} f:name: {} f:value: {} k:{"name":"STORAGE_ALLOW_PATTERNS"}: .: {} f:name: {} f:value: {} f:image: {} f:imagePullPolicy: {} f:name: {} f:resources: .: {} f:limits: .: {} f:cpu: {} f:memory: {} f:requests: .: {} f:cpu: {} f:memory: {} f:terminationMessagePath: {} f:terminationMessagePolicy: {} f:volumeMounts: .: {} k:{"mountPath":"/mnt/models"}: .: {} f:mountPath: {} f:name: {} f:restartPolicy: {} f:schedulerName: {} f:securityContext: {} f:serviceAccount: {} f:serviceAccountName: {} f:terminationGracePeriodSeconds: {} f:volumes: .: {} k:{"name":"kserve-provision-location"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tls-certs"}: .: {} f:name: {} f:secret: .: {} f:defaultMode: {} f:secretName: {} k:{"name":"tokenizer-cache"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-tmp"}: .: {} f:emptyDir: {} f:name: {} k:{"name":"tokenizer-uds"}: .: {} f:emptyDir: {} f:name: {} manager: kube-controller-manager operation: Update time: "2026-07-08T07:55:59Z" - 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: "2026-07-08T07:55:59Z" - 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:initContainerStatuses: {} f:observedGeneration: {} f:phase: {} f:podIP: {} f:podIPs: .: {} k:{"ip":"10.133.0.52"}: .: {} f:ip: {} f:startTime: {} manager: kubelet operation: Update subresource: status time: "2026-07-08T07:56:33Z" name: router-with-refs-test-kserve-router-scheduler-74dd96b744-vhbc5 namespace: kserve-ci-e2e-test ownerReferences: - apiVersion: apps/v1 blockOwnerDeletion: true controller: true kind: ReplicaSet name: router-with-refs-test-kserve-router-scheduler-74dd96b744 uid: 0bbf19de-e126-4d6c-8580-b83de6f3a968 resourceVersion: "66500" uid: 8bfbfa7a-932a-42cf-a283-e8a193a5a671 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 - router-with-refs-test-inference-pool - --pool-namespace - kserve-ci-e2e-test - --zap-encoder - json - --grpc-port - "9002" - --grpc-health-port - "9003" - --enable-cert-reload=true - --secure-serving=true - --model-server-metrics-scheme=https - --cert-path=/var/run/kserve/tls env: - name: SSL_CERT_DIR value: /var/run/kserve/tls:/var/run/secrets/kubernetes.io/serviceaccount:/etc/pki/tls/certs image: ghcr.io/llm-d/llm-d-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-67xld readOnly: true dnsPolicy: ClusterFirst enableServiceLinks: true imagePullSecrets: - name: default-dockercfg-psjvw - name: router-with-refs-test-epp-sa-dockercfg-mwlxf initContainers: - args: - hf://facebook/opt-125m - /mnt/models env: - name: AWS_ACCESS_KEY_ID valueFrom: secretKeyRef: key: AWS_ACCESS_KEY_ID name: seaweedfs-s3-creds - name: AWS_SECRET_ACCESS_KEY valueFrom: secretKeyRef: key: AWS_SECRET_ACCESS_KEY name: seaweedfs-s3-creds - name: S3_USE_HTTPS value: "0" - name: S3_ENDPOINT value: s3-service.kserve:8333 - name: AWS_ENDPOINT_URL value: http://s3-service.kserve:8333 - name: S3_VERIFY_SSL value: "0" - name: AWS_CA_BUNDLE value: /etc/ssl/custom-certs/cabundle.crt - name: AWS_CA_BUNDLE_CONFIGMAP value: odh-kserve-custom-ca-bundle - name: HF_HUB_ENABLE_HF_TRANSFER value: "1" - name: HF_XET_HIGH_PERFORMANCE value: "1" - name: HF_XET_NUM_CONCURRENT_RANGE_GETS value: "8" - name: STORAGE_ALLOW_PATTERNS value: '["tokenizer.json", "tokenizer_config.json", "special_tokens_map.json", "vocab.json", "merges.txt", "config.json", "generation_config.json"]' image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imagePullPolicy: IfNotPresent name: storage-initializer resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi securityContext: allowPrivilegeEscalation: false capabilities: drop: - ALL runAsNonRoot: true runAsUser: 1000690000 terminationMessagePath: /dev/termination-log terminationMessagePolicy: FallbackToLogsOnError volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-67xld readOnly: true nodeName: ip-10-0-136-129.ec2.internal preemptionPolicy: PreemptLowerPriority priority: 0 restartPolicy: Always schedulerName: default-scheduler securityContext: fsGroup: 1000690000 seLinuxOptions: level: s0:c26,c20 seccompProfile: type: RuntimeDefault serviceAccount: router-with-refs-test-epp-sa serviceAccountName: router-with-refs-test-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: router-with-refs-test-kserve-self-signed-certs - emptyDir: {} name: tokenizer-uds - emptyDir: {} name: tokenizer-tmp - emptyDir: {} name: tokenizer-cache - emptyDir: {} name: kserve-provision-location - name: kube-api-access-67xld 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: "2026-07-08T07:56:00Z" observedGeneration: 1 status: "True" type: PodReadyToStartContainers - lastProbeTime: null lastTransitionTime: "2026-07-08T07:56:00Z" observedGeneration: 1 status: "True" type: Initialized - lastProbeTime: null lastTransitionTime: "2026-07-08T07:56:33Z" observedGeneration: 1 status: "True" type: Ready - lastProbeTime: null lastTransitionTime: "2026-07-08T07:56:33Z" observedGeneration: 1 status: "True" type: ContainersReady - lastProbeTime: null lastTransitionTime: "2026-07-08T07:55:59Z" observedGeneration: 1 status: "True" type: PodScheduled containerStatuses: - allocatedResources: cpu: 256m memory: 500Mi containerID: cri-o://6ae10eef43cd8d75e7c789c26525d019c4ac51d5032ef11f0ff673b0d140d4bb 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: "2026-07-08T07:56:01Z" 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-67xld readOnly: true recursiveReadOnly: Disabled hostIP: 10.0.136.129 hostIPs: - ip: 10.0.136.129 initContainerStatuses: - allocatedResources: cpu: 100m memory: 100Mi containerID: cri-o://06c6a2f4f3333c1aec3baf53f6525fe9a18d082756ed5ca194e204e0c139625f image: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 imageID: quay.io/opendatahub/kserve-storage-initializer@sha256:956644def59861b6c4d288d394adc3ba824655b2a247e0388911f52fbf9bf773 lastState: {} name: storage-initializer ready: true resources: limits: cpu: "1" memory: 24Gi requests: cpu: 100m memory: 100Mi restartCount: 0 started: false state: terminated: containerID: cri-o://06c6a2f4f3333c1aec3baf53f6525fe9a18d082756ed5ca194e204e0c139625f exitCode: 0 finishedAt: "2026-07-08T07:56:00Z" reason: Completed startedAt: "2026-07-08T07:55:59Z" user: linux: gid: 0 supplementalGroups: - 0 - 1000690000 uid: 1000690000 volumeMounts: - mountPath: /mnt/models name: kserve-provision-location - mountPath: /var/run/secrets/kubernetes.io/serviceaccount name: kube-api-access-67xld readOnly: true recursiveReadOnly: Disabled observedGeneration: 1 phase: Running podIP: 10.133.0.52 podIPs: - ip: 10.133.0.52 qosClass: Burstable startTime: "2026-07-08T07:55:59Z" kind: PodList metadata: resourceVersion: "98434"