From a14bdcb77c058146c4355386c52c862f94ebc54b Mon Sep 17 00:00:00 2001
From: Trevor Royer <troyer@redhat.com>
Date: Tue, 20 Jan 2026 10:53:42 -0700
Subject: [PATCH 01/12] llm-d wip

---
 charts/llm-d-kserve/.helmignore               |  23 +++
 charts/llm-d-kserve/Chart.yaml                |  24 +++
 charts/llm-d-kserve/templates/NOTES.txt       |  35 ++++
 charts/llm-d-kserve/templates/_helpers.tpl    |  62 +++++++
 charts/llm-d-kserve/templates/deployment.yaml |  78 ++++++++
 charts/llm-d-kserve/templates/hpa.yaml        |  32 ++++
 charts/llm-d-kserve/templates/httproute.yaml  |  38 ++++
 charts/llm-d-kserve/templates/ingress.yaml    |  43 +++++
 .../templates/llminferenceservice.yaml        |  60 +++++++
 charts/llm-d-kserve/templates/service.yaml    |  15 ++
 .../templates/serviceaccount.yaml             |  13 ++
 .../templates/tests/test-connection.yaml      |  15 ++
 charts/llm-d-kserve/values.yaml               | 170 ++++++++++++++++++
 13 files changed, 608 insertions(+)
 create mode 100644 charts/llm-d-kserve/.helmignore
 create mode 100644 charts/llm-d-kserve/Chart.yaml
 create mode 100644 charts/llm-d-kserve/templates/NOTES.txt
 create mode 100644 charts/llm-d-kserve/templates/_helpers.tpl
 create mode 100644 charts/llm-d-kserve/templates/deployment.yaml
 create mode 100644 charts/llm-d-kserve/templates/hpa.yaml
 create mode 100644 charts/llm-d-kserve/templates/httproute.yaml
 create mode 100644 charts/llm-d-kserve/templates/ingress.yaml
 create mode 100644 charts/llm-d-kserve/templates/llminferenceservice.yaml
 create mode 100644 charts/llm-d-kserve/templates/service.yaml
 create mode 100644 charts/llm-d-kserve/templates/serviceaccount.yaml
 create mode 100644 charts/llm-d-kserve/templates/tests/test-connection.yaml
 create mode 100644 charts/llm-d-kserve/values.yaml

diff --git a/charts/llm-d-kserve/.helmignore b/charts/llm-d-kserve/.helmignore
new file mode 100644
index 0000000..0e8a0eb
--- /dev/null
+++ b/charts/llm-d-kserve/.helmignore
@@ -0,0 +1,23 @@
+# Patterns to ignore when building packages.
+# This supports shell glob matching, relative path matching, and
+# negation (prefixed with !). Only one pattern per line.
+.DS_Store
+# Common VCS dirs
+.git/
+.gitignore
+.bzr/
+.bzrignore
+.hg/
+.hgignore
+.svn/
+# Common backup files
+*.swp
+*.bak
+*.tmp
+*.orig
+*~
+# Various IDEs
+.project
+.idea/
+*.tmproj
+.vscode/
diff --git a/charts/llm-d-kserve/Chart.yaml b/charts/llm-d-kserve/Chart.yaml
new file mode 100644
index 0000000..9508087
--- /dev/null
+++ b/charts/llm-d-kserve/Chart.yaml
@@ -0,0 +1,24 @@
+apiVersion: v2
+name: llm-d-kserve
+description: A Helm chart for Kubernetes
+
+# A chart can be either an 'application' or a 'library' chart.
+#
+# Application charts are a collection of templates that can be packaged into versioned archives
+# to be deployed.
+#
+# Library charts provide useful utilities or functions for the chart developer. They're included as
+# a dependency of application charts to inject those utilities and functions into the rendering
+# pipeline. Library charts do not define any templates and therefore cannot be deployed.
+type: application
+
+# This is the chart version. This version number should be incremented each time you make changes
+# to the chart and its templates, including the app version.
+# Versions are expected to follow Semantic Versioning (https://semver.org/)
+version: 0.1.0
+
+# This is the version number of the application being deployed. This version number should be
+# incremented each time you make changes to the application. Versions are not expected to
+# follow Semantic Versioning. They should reflect the version the application is using.
+# It is recommended to use it with quotes.
+appVersion: "1.16.0"
diff --git a/charts/llm-d-kserve/templates/NOTES.txt b/charts/llm-d-kserve/templates/NOTES.txt
new file mode 100644
index 0000000..d2970e2
--- /dev/null
+++ b/charts/llm-d-kserve/templates/NOTES.txt
@@ -0,0 +1,35 @@
+1. Get the application URL by running these commands:
+{{- if .Values.httpRoute.enabled }}
+{{- if .Values.httpRoute.hostnames }}
+    export APP_HOSTNAME={{ .Values.httpRoute.hostnames | first }}
+{{- else }}
+    export APP_HOSTNAME=$(kubectl get --namespace {{(first .Values.httpRoute.parentRefs).namespace | default .Release.Namespace }} gateway/{{ (first .Values.httpRoute.parentRefs).name }} -o jsonpath="{.spec.listeners[0].hostname}")
+  {{- end }}
+{{- if and .Values.httpRoute.rules (first .Values.httpRoute.rules).matches (first (first .Values.httpRoute.rules).matches).path.value }}
+    echo "Visit http://$APP_HOSTNAME{{ (first (first .Values.httpRoute.rules).matches).path.value }} to use your application"
+
+    NOTE: Your HTTPRoute depends on the listener configuration of your gateway and your HTTPRoute rules.
+    The rules can be set for path, method, header and query parameters.
+    You can check the gateway configuration with 'kubectl get --namespace {{(first .Values.httpRoute.parentRefs).namespace | default .Release.Namespace }} gateway/{{ (first .Values.httpRoute.parentRefs).name }} -o yaml'
+{{- end }}
+{{- else if .Values.ingress.enabled }}
+{{- range $host := .Values.ingress.hosts }}
+  {{- range .paths }}
+  http{{ if $.Values.ingress.tls }}s{{ end }}://{{ $host.host }}{{ .path }}
+  {{- end }}
+{{- end }}
+{{- else if contains "NodePort" .Values.service.type }}
+  export NODE_PORT=$(kubectl get --namespace {{ .Release.Namespace }} -o jsonpath="{.spec.ports[0].nodePort}" services {{ include "llm-d-kserve.fullname" . }})
+  export NODE_IP=$(kubectl get nodes --namespace {{ .Release.Namespace }} -o jsonpath="{.items[0].status.addresses[0].address}")
+  echo http://$NODE_IP:$NODE_PORT
+{{- else if contains "LoadBalancer" .Values.service.type }}
+     NOTE: It may take a few minutes for the LoadBalancer IP to be available.
+           You can watch its status by running 'kubectl get --namespace {{ .Release.Namespace }} svc -w {{ include "llm-d-kserve.fullname" . }}'
+  export SERVICE_IP=$(kubectl get svc --namespace {{ .Release.Namespace }} {{ include "llm-d-kserve.fullname" . }} --template "{{"{{ range (index .status.loadBalancer.ingress 0) }}{{.}}{{ end }}"}}")
+  echo http://$SERVICE_IP:{{ .Values.service.port }}
+{{- else if contains "ClusterIP" .Values.service.type }}
+  export POD_NAME=$(kubectl get pods --namespace {{ .Release.Namespace }} -l "app.kubernetes.io/name={{ include "llm-d-kserve.name" . }},app.kubernetes.io/instance={{ .Release.Name }}" -o jsonpath="{.items[0].metadata.name}")
+  export CONTAINER_PORT=$(kubectl get pod --namespace {{ .Release.Namespace }} $POD_NAME -o jsonpath="{.spec.containers[0].ports[0].containerPort}")
+  echo "Visit http://127.0.0.1:8080 to use your application"
+  kubectl --namespace {{ .Release.Namespace }} port-forward $POD_NAME 8080:$CONTAINER_PORT
+{{- end }}
diff --git a/charts/llm-d-kserve/templates/_helpers.tpl b/charts/llm-d-kserve/templates/_helpers.tpl
new file mode 100644
index 0000000..a190bdb
--- /dev/null
+++ b/charts/llm-d-kserve/templates/_helpers.tpl
@@ -0,0 +1,62 @@
+{{/*
+Expand the name of the chart.
+*/}}
+{{- define "llm-d-kserve.name" -}}
+{{- default .Chart.Name .Values.nameOverride | trunc 63 | trimSuffix "-" }}
+{{- end }}
+
+{{/*
+Create a default fully qualified app name.
+We truncate at 63 chars because some Kubernetes name fields are limited to this (by the DNS naming spec).
+If release name contains chart name it will be used as a full name.
+*/}}
+{{- define "llm-d-kserve.fullname" -}}
+{{- if .Values.fullnameOverride }}
+{{- .Values.fullnameOverride | trunc 63 | trimSuffix "-" }}
+{{- else }}
+{{- $name := default .Chart.Name .Values.nameOverride }}
+{{- if contains $name .Release.Name }}
+{{- .Release.Name | trunc 63 | trimSuffix "-" }}
+{{- else }}
+{{- printf "%s-%s" .Release.Name $name | trunc 63 | trimSuffix "-" }}
+{{- end }}
+{{- end }}
+{{- end }}
+
+{{/*
+Create chart name and version as used by the chart label.
+*/}}
+{{- define "llm-d-kserve.chart" -}}
+{{- printf "%s-%s" .Chart.Name .Chart.Version | replace "+" "_" | trunc 63 | trimSuffix "-" }}
+{{- end }}
+
+{{/*
+Common labels
+*/}}
+{{- define "llm-d-kserve.labels" -}}
+helm.sh/chart: {{ include "llm-d-kserve.chart" . }}
+{{ include "llm-d-kserve.selectorLabels" . }}
+{{- if .Chart.AppVersion }}
+app.kubernetes.io/version: {{ .Chart.AppVersion | quote }}
+{{- end }}
+app.kubernetes.io/managed-by: {{ .Release.Service }}
+{{- end }}
+
+{{/*
+Selector labels
+*/}}
+{{- define "llm-d-kserve.selectorLabels" -}}
+app.kubernetes.io/name: {{ include "llm-d-kserve.name" . }}
+app.kubernetes.io/instance: {{ .Release.Name }}
+{{- end }}
+
+{{/*
+Create the name of the service account to use
+*/}}
+{{- define "llm-d-kserve.serviceAccountName" -}}
+{{- if .Values.serviceAccount.create }}
+{{- default (include "llm-d-kserve.fullname" .) .Values.serviceAccount.name }}
+{{- else }}
+{{- default "default" .Values.serviceAccount.name }}
+{{- end }}
+{{- end }}
diff --git a/charts/llm-d-kserve/templates/deployment.yaml b/charts/llm-d-kserve/templates/deployment.yaml
new file mode 100644
index 0000000..1cddb1e
--- /dev/null
+++ b/charts/llm-d-kserve/templates/deployment.yaml
@@ -0,0 +1,78 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: {{ include "llm-d-kserve.fullname" . }}
+  labels:
+    {{- include "llm-d-kserve.labels" . | nindent 4 }}
+spec:
+  {{- if not .Values.autoscaling.enabled }}
+  replicas: {{ .Values.replicaCount }}
+  {{- end }}
+  selector:
+    matchLabels:
+      {{- include "llm-d-kserve.selectorLabels" . | nindent 6 }}
+  template:
+    metadata:
+      {{- with .Values.podAnnotations }}
+      annotations:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      labels:
+        {{- include "llm-d-kserve.labels" . | nindent 8 }}
+        {{- with .Values.podLabels }}
+        {{- toYaml . | nindent 8 }}
+        {{- end }}
+    spec:
+      {{- with .Values.imagePullSecrets }}
+      imagePullSecrets:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      serviceAccountName: {{ include "llm-d-kserve.serviceAccountName" . }}
+      {{- with .Values.podSecurityContext }}
+      securityContext:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      containers:
+        - name: {{ .Chart.Name }}
+          {{- with .Values.securityContext }}
+          securityContext:
+            {{- toYaml . | nindent 12 }}
+          {{- end }}
+          image: "{{ .Values.image.repository }}:{{ .Values.image.tag | default .Chart.AppVersion }}"
+          imagePullPolicy: {{ .Values.image.pullPolicy }}
+          ports:
+            - name: http
+              containerPort: {{ .Values.service.port }}
+              protocol: TCP
+          {{- with .Values.livenessProbe }}
+          livenessProbe:
+            {{- toYaml . | nindent 12 }}
+          {{- end }}
+          {{- with .Values.readinessProbe }}
+          readinessProbe:
+            {{- toYaml . | nindent 12 }}
+          {{- end }}
+          {{- with .Values.resources }}
+          resources:
+            {{- toYaml . | nindent 12 }}
+          {{- end }}
+          {{- with .Values.volumeMounts }}
+          volumeMounts:
+            {{- toYaml . | nindent 12 }}
+          {{- end }}
+      {{- with .Values.volumes }}
+      volumes:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      {{- with .Values.nodeSelector }}
+      nodeSelector:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      {{- with .Values.affinity }}
+      affinity:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      {{- with .Values.tolerations }}
+      tolerations:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
diff --git a/charts/llm-d-kserve/templates/hpa.yaml b/charts/llm-d-kserve/templates/hpa.yaml
new file mode 100644
index 0000000..c5a66f6
--- /dev/null
+++ b/charts/llm-d-kserve/templates/hpa.yaml
@@ -0,0 +1,32 @@
+{{- if .Values.autoscaling.enabled }}
+apiVersion: autoscaling/v2
+kind: HorizontalPodAutoscaler
+metadata:
+  name: {{ include "llm-d-kserve.fullname" . }}
+  labels:
+    {{- include "llm-d-kserve.labels" . | nindent 4 }}
+spec:
+  scaleTargetRef:
+    apiVersion: apps/v1
+    kind: Deployment
+    name: {{ include "llm-d-kserve.fullname" . }}
+  minReplicas: {{ .Values.autoscaling.minReplicas }}
+  maxReplicas: {{ .Values.autoscaling.maxReplicas }}
+  metrics:
+    {{- if .Values.autoscaling.targetCPUUtilizationPercentage }}
+    - type: Resource
+      resource:
+        name: cpu
+        target:
+          type: Utilization
+          averageUtilization: {{ .Values.autoscaling.targetCPUUtilizationPercentage }}
+    {{- end }}
+    {{- if .Values.autoscaling.targetMemoryUtilizationPercentage }}
+    - type: Resource
+      resource:
+        name: memory
+        target:
+          type: Utilization
+          averageUtilization: {{ .Values.autoscaling.targetMemoryUtilizationPercentage }}
+    {{- end }}
+{{- end }}
diff --git a/charts/llm-d-kserve/templates/httproute.yaml b/charts/llm-d-kserve/templates/httproute.yaml
new file mode 100644
index 0000000..0777b66
--- /dev/null
+++ b/charts/llm-d-kserve/templates/httproute.yaml
@@ -0,0 +1,38 @@
+{{- if .Values.httpRoute.enabled -}}
+{{- $fullName := include "llm-d-kserve.fullname" . -}}
+{{- $svcPort := .Values.service.port -}}
+apiVersion: gateway.networking.k8s.io/v1
+kind: HTTPRoute
+metadata:
+  name: {{ $fullName }}
+  labels:
+    {{- include "llm-d-kserve.labels" . | nindent 4 }}
+  {{- with .Values.httpRoute.annotations }}
+  annotations:
+    {{- toYaml . | nindent 4 }}
+  {{- end }}
+spec:
+  parentRefs:
+    {{- with .Values.httpRoute.parentRefs }}
+      {{- toYaml . | nindent 4 }}
+    {{- end }}
+  {{- with .Values.httpRoute.hostnames }}
+  hostnames:
+    {{- toYaml . | nindent 4 }}
+  {{- end }}
+  rules:
+    {{- range .Values.httpRoute.rules }}
+    {{- with .matches }}
+    - matches:
+      {{- toYaml . | nindent 8 }}
+    {{- end }}
+    {{- with .filters }}
+      filters:
+      {{- toYaml . | nindent 8 }}
+    {{- end }}
+      backendRefs:
+        - name: {{ $fullName }}
+          port: {{ $svcPort }}
+          weight: 1
+    {{- end }}
+{{- end }}
diff --git a/charts/llm-d-kserve/templates/ingress.yaml b/charts/llm-d-kserve/templates/ingress.yaml
new file mode 100644
index 0000000..83ecb55
--- /dev/null
+++ b/charts/llm-d-kserve/templates/ingress.yaml
@@ -0,0 +1,43 @@
+{{- if .Values.ingress.enabled -}}
+apiVersion: networking.k8s.io/v1
+kind: Ingress
+metadata:
+  name: {{ include "llm-d-kserve.fullname" . }}
+  labels:
+    {{- include "llm-d-kserve.labels" . | nindent 4 }}
+  {{- with .Values.ingress.annotations }}
+  annotations:
+    {{- toYaml . | nindent 4 }}
+  {{- end }}
+spec:
+  {{- with .Values.ingress.className }}
+  ingressClassName: {{ . }}
+  {{- end }}
+  {{- if .Values.ingress.tls }}
+  tls:
+    {{- range .Values.ingress.tls }}
+    - hosts:
+        {{- range .hosts }}
+        - {{ . | quote }}
+        {{- end }}
+      secretName: {{ .secretName }}
+    {{- end }}
+  {{- end }}
+  rules:
+    {{- range .Values.ingress.hosts }}
+    - host: {{ .host | quote }}
+      http:
+        paths:
+          {{- range .paths }}
+          - path: {{ .path }}
+            {{- with .pathType }}
+            pathType: {{ . }}
+            {{- end }}
+            backend:
+              service:
+                name: {{ include "llm-d-kserve.fullname" $ }}
+                port:
+                  number: {{ $.Values.service.port }}
+          {{- end }}
+    {{- end }}
+{{- end }}
diff --git a/charts/llm-d-kserve/templates/llminferenceservice.yaml b/charts/llm-d-kserve/templates/llminferenceservice.yaml
new file mode 100644
index 0000000..d17b40e
--- /dev/null
+++ b/charts/llm-d-kserve/templates/llminferenceservice.yaml
@@ -0,0 +1,60 @@
+apiVersion: serving.kserve.io/v1alpha1
+kind: LLMInferenceService
+metadata:
+  name: {{ include "llm-d-kserve.fullname" . }}
+  labels:
+    {{- include "llm-d-kserve.labels" . | nindent 4 }}
+spec:
+  model:
+    name: {{ .Values.model.nameOverride | default (include "llm-d-kserve.fullname" .) }}
+    uri: {{ .Values.model.uri }}
+  replicas: {{ .Values.replicaCount }}
+  router:
+    scheduler:
+      template:
+        containers:
+          - name: main
+            args:
+              - -v=4
+              - --poolName
+              - '{{ ChildName .ObjectMeta.Name `-inference-pool` }}'
+              - --poolNamespace
+              - '{{ .ObjectMeta.Namespace }}'
+              - --zap-encoder
+              - json
+              - --grpcPort
+              - "9002"
+              - --grpcHealthPort
+              - "9003"
+              - --secureServing
+              - --modelServerMetricsScheme
+              - https
+              - --modelServerMetricsHttpsInsecureSkipVerify
+              - --certPath
+              - /etc/ssl/certs
+              - --configText
+              - |2
+
+                apiVersion: inference.networking.x-k8s.io/v1alpha1
+                kind: EndpointPickerConfig
+                plugins:
+                  - type: single-profile-handler
+                  - type: max-score-picker
+                  - type: queue-scorer
+                  - type: kv-cache-scorer
+                  - type: prefix-cache-scorer
+                    parameters:
+                      hashBlockSize: 64             # block size 16 * 4 chars / token
+                      maxPrefixBlocksToMatch: 256   # this matches up to ~4k tokens
+                      lruCapacityPerServer: 39744   # total blocks (635904) / block size (16)
+                schedulingProfiles:
+                  - name: default
+                    plugins:
+                      - pluginRef: prefix-cache-scorer
+                        weight: 2.0
+                      - pluginRef: queue-scorer
+                        weight: 1.0
+                      - pluginRef: kv-cache-scorer
+                        weight: 1.0
+                      - pluginRef: max-score-picker
+            
\ No newline at end of file
diff --git a/charts/llm-d-kserve/templates/service.yaml b/charts/llm-d-kserve/templates/service.yaml
new file mode 100644
index 0000000..0705876
--- /dev/null
+++ b/charts/llm-d-kserve/templates/service.yaml
@@ -0,0 +1,15 @@
+apiVersion: v1
+kind: Service
+metadata:
+  name: {{ include "llm-d-kserve.fullname" . }}
+  labels:
+    {{- include "llm-d-kserve.labels" . | nindent 4 }}
+spec:
+  type: {{ .Values.service.type }}
+  ports:
+    - port: {{ .Values.service.port }}
+      targetPort: http
+      protocol: TCP
+      name: http
+  selector:
+    {{- include "llm-d-kserve.selectorLabels" . | nindent 4 }}
diff --git a/charts/llm-d-kserve/templates/serviceaccount.yaml b/charts/llm-d-kserve/templates/serviceaccount.yaml
new file mode 100644
index 0000000..39574b5
--- /dev/null
+++ b/charts/llm-d-kserve/templates/serviceaccount.yaml
@@ -0,0 +1,13 @@
+{{- if .Values.serviceAccount.create -}}
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: {{ include "llm-d-kserve.serviceAccountName" . }}
+  labels:
+    {{- include "llm-d-kserve.labels" . | nindent 4 }}
+  {{- with .Values.serviceAccount.annotations }}
+  annotations:
+    {{- toYaml . | nindent 4 }}
+  {{- end }}
+automountServiceAccountToken: {{ .Values.serviceAccount.automount }}
+{{- end }}
diff --git a/charts/llm-d-kserve/templates/tests/test-connection.yaml b/charts/llm-d-kserve/templates/tests/test-connection.yaml
new file mode 100644
index 0000000..cd3bc3f
--- /dev/null
+++ b/charts/llm-d-kserve/templates/tests/test-connection.yaml
@@ -0,0 +1,15 @@
+apiVersion: v1
+kind: Pod
+metadata:
+  name: "{{ include "llm-d-kserve.fullname" . }}-test-connection"
+  labels:
+    {{- include "llm-d-kserve.labels" . | nindent 4 }}
+  annotations:
+    "helm.sh/hook": test
+spec:
+  containers:
+    - name: wget
+      image: busybox
+      command: ['wget']
+      args: ['{{ include "llm-d-kserve.fullname" . }}:{{ .Values.service.port }}']
+  restartPolicy: Never
diff --git a/charts/llm-d-kserve/values.yaml b/charts/llm-d-kserve/values.yaml
new file mode 100644
index 0000000..91faf6a
--- /dev/null
+++ b/charts/llm-d-kserve/values.yaml
@@ -0,0 +1,170 @@
+# Default values for llm-d-kserve.
+# This is a YAML-formatted file.
+# Declare variables to be passed into your templates.
+
+# This will set the replicaset count more information can be found here: https://kubernetes.io/docs/concepts/workloads/controllers/replicaset/
+replicaCount: 1
+
+model:
+  nameOverride: "meta-llama/llama-3.2-1b-instruct"
+  uri: oci://quay.io/redhat-ai-services/modelcar-catalog:llama-3.2-1b-instruct
+
+tolerations:
+  - key: "nvidia.com/gpu"
+    operator: "Exists"
+    effect: "NoSchedule"
+
+# This sets the container image more information can be found here: https://kubernetes.io/docs/concepts/containers/images/
+image:
+  repository: nginx
+  # This sets the pull policy for images.
+  pullPolicy: IfNotPresent
+  # Overrides the image tag whose default is the chart appVersion.
+  tag: ""
+
+# This is for the secrets for pulling an image from a private repository more information can be found here: https://kubernetes.io/docs/tasks/configure-pod-container/pull-image-private-registry/
+imagePullSecrets: []
+# This is to override the chart name.
+nameOverride: ""
+fullnameOverride: ""
+
+# This section builds out the service account more information can be found here: https://kubernetes.io/docs/concepts/security/service-accounts/
+serviceAccount:
+  # Specifies whether a service account should be created
+  create: true
+  # Automatically mount a ServiceAccount's API credentials?
+  automount: true
+  # Annotations to add to the service account
+  annotations: {}
+  # The name of the service account to use.
+  # If not set and create is true, a name is generated using the fullname template
+  name: ""
+
+# This is for setting Kubernetes Annotations to a Pod.
+# For more information checkout: https://kubernetes.io/docs/concepts/overview/working-with-objects/annotations/
+podAnnotations: {}
+# This is for setting Kubernetes Labels to a Pod.
+# For more information checkout: https://kubernetes.io/docs/concepts/overview/working-with-objects/labels/
+podLabels: {}
+
+podSecurityContext: {}
+  # fsGroup: 2000
+
+securityContext: {}
+  # capabilities:
+  #   drop:
+  #   - ALL
+  # readOnlyRootFilesystem: true
+  # runAsNonRoot: true
+  # runAsUser: 1000
+
+# This is for setting up a service more information can be found here: https://kubernetes.io/docs/concepts/services-networking/service/
+service:
+  # This sets the service type more information can be found here: https://kubernetes.io/docs/concepts/services-networking/service/#publishing-services-service-types
+  type: ClusterIP
+  # This sets the ports more information can be found here: https://kubernetes.io/docs/concepts/services-networking/service/#field-spec-ports
+  port: 80
+
+# This block is for setting up the ingress for more information can be found here: https://kubernetes.io/docs/concepts/services-networking/ingress/
+ingress:
+  enabled: false
+  className: ""
+  annotations: {}
+    # kubernetes.io/ingress.class: nginx
+    # kubernetes.io/tls-acme: "true"
+  hosts:
+    - host: chart-example.local
+      paths:
+        - path: /
+          pathType: ImplementationSpecific
+  tls: []
+  #  - secretName: chart-example-tls
+  #    hosts:
+  #      - chart-example.local
+
+# -- Expose the service via gateway-api HTTPRoute
+# Requires Gateway API resources and suitable controller installed within the cluster
+# (see: https://gateway-api.sigs.k8s.io/guides/)
+httpRoute:
+  # HTTPRoute enabled.
+  enabled: false
+  # HTTPRoute annotations.
+  annotations: {}
+  # Which Gateways this Route is attached to.
+  parentRefs:
+  - name: gateway
+    sectionName: http
+    # namespace: default
+  # Hostnames matching HTTP header.
+  hostnames:
+  - chart-example.local
+  # List of rules and filters applied.
+  rules:
+  - matches:
+    - path:
+        type: PathPrefix
+        value: /headers
+  #   filters:
+  #   - type: RequestHeaderModifier
+  #     requestHeaderModifier:
+  #       set:
+  #       - name: My-Overwrite-Header
+  #         value: this-is-the-only-value
+  #       remove:
+  #       - User-Agent
+  # - matches:
+  #   - path:
+  #       type: PathPrefix
+  #       value: /echo
+  #     headers:
+  #     - name: version
+  #       value: v2
+
+resources: {}
+  # We usually recommend not to specify default resources and to leave this as a conscious
+  # choice for the user. This also increases chances charts run on environments with little
+  # resources, such as Minikube. If you do want to specify resources, uncomment the following
+  # lines, adjust them as necessary, and remove the curly braces after 'resources:'.
+  # limits:
+  #   cpu: 100m
+  #   memory: 128Mi
+  # requests:
+  #   cpu: 100m
+  #   memory: 128Mi
+
+# This is to setup the liveness and readiness probes more information can be found here: https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/
+livenessProbe:
+  httpGet:
+    path: /
+    port: http
+readinessProbe:
+  httpGet:
+    path: /
+    port: http
+
+# This section is for setting up autoscaling more information can be found here: https://kubernetes.io/docs/concepts/workloads/autoscaling/
+autoscaling:
+  enabled: false
+  minReplicas: 1
+  maxReplicas: 100
+  targetCPUUtilizationPercentage: 80
+  # targetMemoryUtilizationPercentage: 80
+
+# Additional volumes on the output Deployment definition.
+volumes: []
+# - name: foo
+#   secret:
+#     secretName: mysecret
+#     optional: false
+
+# Additional volumeMounts on the output Deployment definition.
+volumeMounts: []
+# - name: foo
+#   mountPath: "/etc/foo"
+#   readOnly: true
+
+nodeSelector: {}
+
+tolerations: []
+
+affinity: {}

From 6a5618b165b83e756a68ed7bec24a5482574efe6 Mon Sep 17 00:00:00 2001
From: Trevor Royer <troyer@redhat.com>
Date: Fri, 13 Mar 2026 15:37:15 -0500
Subject: [PATCH 02/12] remove unused files

---
 charts/llm-d-kserve/templates/deployment.yaml | 78 -------------------
 charts/llm-d-kserve/templates/hpa.yaml        | 32 --------
 charts/llm-d-kserve/templates/httproute.yaml  | 38 ---------
 charts/llm-d-kserve/templates/ingress.yaml    | 43 ----------
 charts/llm-d-kserve/templates/service.yaml    | 15 ----
 .../templates/tests/test-connection.yaml      | 15 ----
 6 files changed, 221 deletions(-)
 delete mode 100644 charts/llm-d-kserve/templates/deployment.yaml
 delete mode 100644 charts/llm-d-kserve/templates/hpa.yaml
 delete mode 100644 charts/llm-d-kserve/templates/httproute.yaml
 delete mode 100644 charts/llm-d-kserve/templates/ingress.yaml
 delete mode 100644 charts/llm-d-kserve/templates/service.yaml
 delete mode 100644 charts/llm-d-kserve/templates/tests/test-connection.yaml

diff --git a/charts/llm-d-kserve/templates/deployment.yaml b/charts/llm-d-kserve/templates/deployment.yaml
deleted file mode 100644
index 1cddb1e..0000000
--- a/charts/llm-d-kserve/templates/deployment.yaml
+++ /dev/null
@@ -1,78 +0,0 @@
-apiVersion: apps/v1
-kind: Deployment
-metadata:
-  name: {{ include "llm-d-kserve.fullname" . }}
-  labels:
-    {{- include "llm-d-kserve.labels" . | nindent 4 }}
-spec:
-  {{- if not .Values.autoscaling.enabled }}
-  replicas: {{ .Values.replicaCount }}
-  {{- end }}
-  selector:
-    matchLabels:
-      {{- include "llm-d-kserve.selectorLabels" . | nindent 6 }}
-  template:
-    metadata:
-      {{- with .Values.podAnnotations }}
-      annotations:
-        {{- toYaml . | nindent 8 }}
-      {{- end }}
-      labels:
-        {{- include "llm-d-kserve.labels" . | nindent 8 }}
-        {{- with .Values.podLabels }}
-        {{- toYaml . | nindent 8 }}
-        {{- end }}
-    spec:
-      {{- with .Values.imagePullSecrets }}
-      imagePullSecrets:
-        {{- toYaml . | nindent 8 }}
-      {{- end }}
-      serviceAccountName: {{ include "llm-d-kserve.serviceAccountName" . }}
-      {{- with .Values.podSecurityContext }}
-      securityContext:
-        {{- toYaml . | nindent 8 }}
-      {{- end }}
-      containers:
-        - name: {{ .Chart.Name }}
-          {{- with .Values.securityContext }}
-          securityContext:
-            {{- toYaml . | nindent 12 }}
-          {{- end }}
-          image: "{{ .Values.image.repository }}:{{ .Values.image.tag | default .Chart.AppVersion }}"
-          imagePullPolicy: {{ .Values.image.pullPolicy }}
-          ports:
-            - name: http
-              containerPort: {{ .Values.service.port }}
-              protocol: TCP
-          {{- with .Values.livenessProbe }}
-          livenessProbe:
-            {{- toYaml . | nindent 12 }}
-          {{- end }}
-          {{- with .Values.readinessProbe }}
-          readinessProbe:
-            {{- toYaml . | nindent 12 }}
-          {{- end }}
-          {{- with .Values.resources }}
-          resources:
-            {{- toYaml . | nindent 12 }}
-          {{- end }}
-          {{- with .Values.volumeMounts }}
-          volumeMounts:
-            {{- toYaml . | nindent 12 }}
-          {{- end }}
-      {{- with .Values.volumes }}
-      volumes:
-        {{- toYaml . | nindent 8 }}
-      {{- end }}
-      {{- with .Values.nodeSelector }}
-      nodeSelector:
-        {{- toYaml . | nindent 8 }}
-      {{- end }}
-      {{- with .Values.affinity }}
-      affinity:
-        {{- toYaml . | nindent 8 }}
-      {{- end }}
-      {{- with .Values.tolerations }}
-      tolerations:
-        {{- toYaml . | nindent 8 }}
-      {{- end }}
diff --git a/charts/llm-d-kserve/templates/hpa.yaml b/charts/llm-d-kserve/templates/hpa.yaml
deleted file mode 100644
index c5a66f6..0000000
--- a/charts/llm-d-kserve/templates/hpa.yaml
+++ /dev/null
@@ -1,32 +0,0 @@
-{{- if .Values.autoscaling.enabled }}
-apiVersion: autoscaling/v2
-kind: HorizontalPodAutoscaler
-metadata:
-  name: {{ include "llm-d-kserve.fullname" . }}
-  labels:
-    {{- include "llm-d-kserve.labels" . | nindent 4 }}
-spec:
-  scaleTargetRef:
-    apiVersion: apps/v1
-    kind: Deployment
-    name: {{ include "llm-d-kserve.fullname" . }}
-  minReplicas: {{ .Values.autoscaling.minReplicas }}
-  maxReplicas: {{ .Values.autoscaling.maxReplicas }}
-  metrics:
-    {{- if .Values.autoscaling.targetCPUUtilizationPercentage }}
-    - type: Resource
-      resource:
-        name: cpu
-        target:
-          type: Utilization
-          averageUtilization: {{ .Values.autoscaling.targetCPUUtilizationPercentage }}
-    {{- end }}
-    {{- if .Values.autoscaling.targetMemoryUtilizationPercentage }}
-    - type: Resource
-      resource:
-        name: memory
-        target:
-          type: Utilization
-          averageUtilization: {{ .Values.autoscaling.targetMemoryUtilizationPercentage }}
-    {{- end }}
-{{- end }}
diff --git a/charts/llm-d-kserve/templates/httproute.yaml b/charts/llm-d-kserve/templates/httproute.yaml
deleted file mode 100644
index 0777b66..0000000
--- a/charts/llm-d-kserve/templates/httproute.yaml
+++ /dev/null
@@ -1,38 +0,0 @@
-{{- if .Values.httpRoute.enabled -}}
-{{- $fullName := include "llm-d-kserve.fullname" . -}}
-{{- $svcPort := .Values.service.port -}}
-apiVersion: gateway.networking.k8s.io/v1
-kind: HTTPRoute
-metadata:
-  name: {{ $fullName }}
-  labels:
-    {{- include "llm-d-kserve.labels" . | nindent 4 }}
-  {{- with .Values.httpRoute.annotations }}
-  annotations:
-    {{- toYaml . | nindent 4 }}
-  {{- end }}
-spec:
-  parentRefs:
-    {{- with .Values.httpRoute.parentRefs }}
-      {{- toYaml . | nindent 4 }}
-    {{- end }}
-  {{- with .Values.httpRoute.hostnames }}
-  hostnames:
-    {{- toYaml . | nindent 4 }}
-  {{- end }}
-  rules:
-    {{- range .Values.httpRoute.rules }}
-    {{- with .matches }}
-    - matches:
-      {{- toYaml . | nindent 8 }}
-    {{- end }}
-    {{- with .filters }}
-      filters:
-      {{- toYaml . | nindent 8 }}
-    {{- end }}
-      backendRefs:
-        - name: {{ $fullName }}
-          port: {{ $svcPort }}
-          weight: 1
-    {{- end }}
-{{- end }}
diff --git a/charts/llm-d-kserve/templates/ingress.yaml b/charts/llm-d-kserve/templates/ingress.yaml
deleted file mode 100644
index 83ecb55..0000000
--- a/charts/llm-d-kserve/templates/ingress.yaml
+++ /dev/null
@@ -1,43 +0,0 @@
-{{- if .Values.ingress.enabled -}}
-apiVersion: networking.k8s.io/v1
-kind: Ingress
-metadata:
-  name: {{ include "llm-d-kserve.fullname" . }}
-  labels:
-    {{- include "llm-d-kserve.labels" . | nindent 4 }}
-  {{- with .Values.ingress.annotations }}
-  annotations:
-    {{- toYaml . | nindent 4 }}
-  {{- end }}
-spec:
-  {{- with .Values.ingress.className }}
-  ingressClassName: {{ . }}
-  {{- end }}
-  {{- if .Values.ingress.tls }}
-  tls:
-    {{- range .Values.ingress.tls }}
-    - hosts:
-        {{- range .hosts }}
-        - {{ . | quote }}
-        {{- end }}
-      secretName: {{ .secretName }}
-    {{- end }}
-  {{- end }}
-  rules:
-    {{- range .Values.ingress.hosts }}
-    - host: {{ .host | quote }}
-      http:
-        paths:
-          {{- range .paths }}
-          - path: {{ .path }}
-            {{- with .pathType }}
-            pathType: {{ . }}
-            {{- end }}
-            backend:
-              service:
-                name: {{ include "llm-d-kserve.fullname" $ }}
-                port:
-                  number: {{ $.Values.service.port }}
-          {{- end }}
-    {{- end }}
-{{- end }}
diff --git a/charts/llm-d-kserve/templates/service.yaml b/charts/llm-d-kserve/templates/service.yaml
deleted file mode 100644
index 0705876..0000000
--- a/charts/llm-d-kserve/templates/service.yaml
+++ /dev/null
@@ -1,15 +0,0 @@
-apiVersion: v1
-kind: Service
-metadata:
-  name: {{ include "llm-d-kserve.fullname" . }}
-  labels:
-    {{- include "llm-d-kserve.labels" . | nindent 4 }}
-spec:
-  type: {{ .Values.service.type }}
-  ports:
-    - port: {{ .Values.service.port }}
-      targetPort: http
-      protocol: TCP
-      name: http
-  selector:
-    {{- include "llm-d-kserve.selectorLabels" . | nindent 4 }}
diff --git a/charts/llm-d-kserve/templates/tests/test-connection.yaml b/charts/llm-d-kserve/templates/tests/test-connection.yaml
deleted file mode 100644
index cd3bc3f..0000000
--- a/charts/llm-d-kserve/templates/tests/test-connection.yaml
+++ /dev/null
@@ -1,15 +0,0 @@
-apiVersion: v1
-kind: Pod
-metadata:
-  name: "{{ include "llm-d-kserve.fullname" . }}-test-connection"
-  labels:
-    {{- include "llm-d-kserve.labels" . | nindent 4 }}
-  annotations:
-    "helm.sh/hook": test
-spec:
-  containers:
-    - name: wget
-      image: busybox
-      command: ['wget']
-      args: ['{{ include "llm-d-kserve.fullname" . }}:{{ .Values.service.port }}']
-  restartPolicy: Never

From 3d3ca3468c7c78b1bbdd0459c0f6916753fd681b Mon Sep 17 00:00:00 2001
From: Trevor Royer <troyer@redhat.com>
Date: Fri, 13 Mar 2026 15:40:26 -0500
Subject: [PATCH 03/12] update helm chart with new values config

---
 charts/llm-d-kserve/Chart.yaml                |  28 +-
 charts/llm-d-kserve/README.md.gotmpl          |  50 ++++
 charts/llm-d-kserve/templates/NOTES.txt       |  96 ++++---
 charts/llm-d-kserve/templates/_helpers.tpl    |   8 +
 .../templates/hardwareprofile.yaml            |  21 ++
 .../templates/llminferenceservice.yaml        |  94 +++----
 charts/llm-d-kserve/values.yaml               | 259 +++++++-----------
 7 files changed, 293 insertions(+), 263 deletions(-)
 create mode 100644 charts/llm-d-kserve/README.md.gotmpl
 create mode 100644 charts/llm-d-kserve/templates/hardwareprofile.yaml

diff --git a/charts/llm-d-kserve/Chart.yaml b/charts/llm-d-kserve/Chart.yaml
index 9508087..2c6e0f7 100644
--- a/charts/llm-d-kserve/Chart.yaml
+++ b/charts/llm-d-kserve/Chart.yaml
@@ -1,24 +1,14 @@
 apiVersion: v2
-name: llm-d-kserve
-description: A Helm chart for Kubernetes
-
-# A chart can be either an 'application' or a 'library' chart.
-#
-# Application charts are a collection of templates that can be packaged into versioned archives
-# to be deployed.
-#
-# Library charts provide useful utilities or functions for the chart developer. They're included as
-# a dependency of application charts to inject those utilities and functions into the rendering
-# pipeline. Library charts do not define any templates and therefore cannot be deployed.
 type: application
 
-# This is the chart version. This version number should be incremented each time you make changes
-# to the chart and its templates, including the app version.
-# Versions are expected to follow Semantic Versioning (https://semver.org/)
+name: llm-d-kserve
+description: A Helm chart for deploying LLM-D with KServe on OpenShift AI
+
 version: 0.1.0
 
-# This is the version number of the application being deployed. This version number should be
-# incremented each time you make changes to the application. Versions are not expected to
-# follow Semantic Versioning. They should reflect the version the application is using.
-# It is recommended to use it with quotes.
-appVersion: "1.16.0"
+# OpenShift AI version
+appVersion: "3.3.0"
+
+icon: https://avatars.githubusercontent.com/u/211385051?s=48&v=4
+sources:
+  - https://github.com/redhat-ai-services/helm-charts/tree/main/charts/llm-d-kserve
diff --git a/charts/llm-d-kserve/README.md.gotmpl b/charts/llm-d-kserve/README.md.gotmpl
new file mode 100644
index 0000000..28ca528
--- /dev/null
+++ b/charts/llm-d-kserve/README.md.gotmpl
@@ -0,0 +1,50 @@
+{{ template "chart.header" . }}
+{{ template "chart.description" . }}
+
+{{ template "chart.versionBadge" . }}{{ template "chart.typeBadge" . }}{{ template "chart.appVersionBadge" . }}
+
+## Installing the Chart
+
+To access charts from this from the cli repository add it:
+
+```sh
+helm repo add redhat-ai-services https://redhat-ai-services.github.io/helm-charts/
+helm repo update redhat-ai-services
+helm upgrade -i [release-name] redhat-ai-services/{{ template "chart.name" . }}
+```
+
+To include a chart from this repository in an umbrella chart, include it in your dependencies in your `Chart.yaml` file.
+
+```yaml
+apiVersion: v2
+name: example-chart
+description: A Helm chart for Kubernetes
+type: application
+
+version: 0.1.0
+
+appVersion: "1.16.0"
+
+dependencies:
+  - name: "{{ template "chart.name" . }}"
+    version: "{{ template "chart.version" . }}"
+    repository: "https://redhat-ai-services.github.io/helm-charts/"
+```
+
+## Usage
+
+### Quick Start
+
+TODO
+
+## Configuration
+
+For a complete list of all configuration options, see the [Values](#values) section below.
+
+{{ template "chart.sourcesSection" . }}
+
+{{ template "chart.requirementsSection" . }}
+
+{{ template "chart.valuesSection" . }}
+
+{{ template "helm-docs.versionFooter" . }}
diff --git a/charts/llm-d-kserve/templates/NOTES.txt b/charts/llm-d-kserve/templates/NOTES.txt
index d2970e2..3a72982 100644
--- a/charts/llm-d-kserve/templates/NOTES.txt
+++ b/charts/llm-d-kserve/templates/NOTES.txt
@@ -1,35 +1,65 @@
-1. Get the application URL by running these commands:
-{{- if .Values.httpRoute.enabled }}
-{{- if .Values.httpRoute.hostnames }}
-    export APP_HOSTNAME={{ .Values.httpRoute.hostnames | first }}
+Successfully deployed LLM-D InferenceService!
+
+DEPLOYMENT INFORMATION:
+  Name: {{ include "llm-d-kserve.fullname" . }}
+  Namespace: {{ .Release.Namespace }}
+  Model: {{ .Values.model.name }}
+  Model URI: {{ .Values.model.uri }}
+  Replicas: {{ .Values.vllm.replicaCount }}
+{{- if .Values.hardwareProfile.create }}
+
+  A hardware profile "{{ include "llm-d-kserve.fullname" . }}" has been created in this namespace.
 {{- else }}
-    export APP_HOSTNAME=$(kubectl get --namespace {{(first .Values.httpRoute.parentRefs).namespace | default .Release.Namespace }} gateway/{{ (first .Values.httpRoute.parentRefs).name }} -o jsonpath="{.spec.listeners[0].hostname}")
-  {{- end }}
-{{- if and .Values.httpRoute.rules (first .Values.httpRoute.rules).matches (first (first .Values.httpRoute.rules).matches).path.value }}
-    echo "Visit http://$APP_HOSTNAME{{ (first (first .Values.httpRoute.rules).matches).path.value }} to use your application"
-
-    NOTE: Your HTTPRoute depends on the listener configuration of your gateway and your HTTPRoute rules.
-    The rules can be set for path, method, header and query parameters.
-    You can check the gateway configuration with 'kubectl get --namespace {{(first .Values.httpRoute.parentRefs).namespace | default .Release.Namespace }} gateway/{{ (first .Values.httpRoute.parentRefs).name }} -o yaml'
-{{- end }}
-{{- else if .Values.ingress.enabled }}
-{{- range $host := .Values.ingress.hosts }}
-  {{- range .paths }}
-  http{{ if $.Values.ingress.tls }}s{{ end }}://{{ $host.host }}{{ .path }}
-  {{- end }}
-{{- end }}
-{{- else if contains "NodePort" .Values.service.type }}
-  export NODE_PORT=$(kubectl get --namespace {{ .Release.Namespace }} -o jsonpath="{.spec.ports[0].nodePort}" services {{ include "llm-d-kserve.fullname" . }})
-  export NODE_IP=$(kubectl get nodes --namespace {{ .Release.Namespace }} -o jsonpath="{.items[0].status.addresses[0].address}")
-  echo http://$NODE_IP:$NODE_PORT
-{{- else if contains "LoadBalancer" .Values.service.type }}
-     NOTE: It may take a few minutes for the LoadBalancer IP to be available.
-           You can watch its status by running 'kubectl get --namespace {{ .Release.Namespace }} svc -w {{ include "llm-d-kserve.fullname" . }}'
-  export SERVICE_IP=$(kubectl get svc --namespace {{ .Release.Namespace }} {{ include "llm-d-kserve.fullname" . }} --template "{{"{{ range (index .status.loadBalancer.ingress 0) }}{{.}}{{ end }}"}}")
-  echo http://$SERVICE_IP:{{ .Values.service.port }}
-{{- else if contains "ClusterIP" .Values.service.type }}
-  export POD_NAME=$(kubectl get pods --namespace {{ .Release.Namespace }} -l "app.kubernetes.io/name={{ include "llm-d-kserve.name" . }},app.kubernetes.io/instance={{ .Release.Name }}" -o jsonpath="{.items[0].metadata.name}")
-  export CONTAINER_PORT=$(kubectl get pod --namespace {{ .Release.Namespace }} $POD_NAME -o jsonpath="{.spec.containers[0].ports[0].containerPort}")
-  echo "Visit http://127.0.0.1:8080 to use your application"
-  kubectl --namespace {{ .Release.Namespace }} port-forward $POD_NAME 8080:$CONTAINER_PORT
+
+  Hardware Profile: {{ .Values.hardwareProfile.name }} (namespace: {{ .Values.hardwareProfile.namespace }})
 {{- end }}
+
+CHECK STATUS:
+  oc get llminferenceservices {{ include "llm-d-kserve.fullname" . }} -n {{ .Release.Namespace }}
+  oc get pods -l serving.kserve.io/llminferenceservice={{ include "llm-d-kserve.fullname" . }} -n {{ .Release.Namespace }}
+
+INTERNAL ACCESS:
+  Your model is configured for internal cluster access.
+
+  To get the service URL (if available):
+  oc get llminferenceservices {{ include "llm-d-kserve.fullname" . }} -n {{ .Release.Namespace }} -o jsonpath='{.status.url}'
+
+  For local access, use port forwarding:
+  oc port-forward svc/{{ include "llm-d-kserve.fullname" . }}-predictor -n {{ .Release.Namespace }} 8080:80
+
+  Then test with:
+  curl -X POST http://localhost:8080/v1/chat/completions \
+    -H "Content-Type: application/json" \
+    -d '{
+      "model": "{{ .Values.model.name }}",
+      "messages": [{"role": "user", "content": "Hello, how are you?"}],
+      "max_tokens": 50
+    }'
+
+  Or for completions:
+  curl -X POST http://localhost:8080/v1/completions \
+    -H "Content-Type: application/json" \
+    -d '{
+      "model": "{{ .Values.model.name }}",
+      "prompt": "Hello, how are you?",
+      "max_tokens": 50
+    }'
+
+USEFUL COMMANDS:
+  # Check pod logs
+  oc logs -l serving.kserve.io/llminferenceservice={{ include "llm-d-kserve.fullname" . }} -n {{ .Release.Namespace }} -f
+
+  # List available models
+  curl http://localhost:8080/v1/models
+
+  # Health check
+  curl http://localhost:8080/health
+
+API DOCUMENTATION:
+  LLM-D exposes an OpenAI-compatible API. Available endpoints:
+  - GET  /v1/models           - List available models
+  - POST /v1/chat/completions - Chat completion
+  - POST /v1/completions      - Text completion
+  - GET  /health              - Health check
+
+  Use port-forward (see above) before calling these endpoints from your machine.
diff --git a/charts/llm-d-kserve/templates/_helpers.tpl b/charts/llm-d-kserve/templates/_helpers.tpl
index a190bdb..db28079 100644
--- a/charts/llm-d-kserve/templates/_helpers.tpl
+++ b/charts/llm-d-kserve/templates/_helpers.tpl
@@ -60,3 +60,11 @@ Create the name of the service account to use
 {{- default "default" .Values.serviceAccount.name }}
 {{- end }}
 {{- end }}
+
+{{- define "llm-d-kserve.image" -}}
+{{- if hasPrefix "sha256:" (toString .Values.vllm.image.tag) }}
+{{- printf "%s@%s" .Values.vllm.image.repository .Values.vllm.image.tag }}
+{{- else }}
+{{- printf "%s:%s" .Values.vllm.image.repository .Values.vllm.image.tag }}
+{{- end }}
+{{- end }}
diff --git a/charts/llm-d-kserve/templates/hardwareprofile.yaml b/charts/llm-d-kserve/templates/hardwareprofile.yaml
new file mode 100644
index 0000000..ba57b5e
--- /dev/null
+++ b/charts/llm-d-kserve/templates/hardwareprofile.yaml
@@ -0,0 +1,21 @@
+{{- if .Values.hardwareProfile.create }}
+apiVersion: infrastructure.opendatahub.io/v1
+kind: HardwareProfile
+metadata:
+  annotations:
+    opendatahub.io/dashboard-feature-visibility: '["model-serving"]'
+    opendatahub.io/description: ""
+    opendatahub.io/disabled: "false"
+    opendatahub.io/display-name: {{ include "llm-d-kserve.fullname" . | quote }}
+  name: {{ include "llm-d-kserve.fullname" . }}
+spec:
+  identifiers:
+    {{- toYaml .Values.hardwareProfile.identifiers | nindent 4 }}
+  scheduling:
+    {{- if .Values.hardwareProfile.tolerations }}
+    node:
+      tolerations:
+        {{- toYaml .Values.hardwareProfile.tolerations | nindent 8 }}
+    {{- end }}
+    type: Node
+{{- end }}
diff --git a/charts/llm-d-kserve/templates/llminferenceservice.yaml b/charts/llm-d-kserve/templates/llminferenceservice.yaml
index d17b40e..bff42be 100644
--- a/charts/llm-d-kserve/templates/llminferenceservice.yaml
+++ b/charts/llm-d-kserve/templates/llminferenceservice.yaml
@@ -1,60 +1,48 @@
 apiVersion: serving.kserve.io/v1alpha1
 kind: LLMInferenceService
 metadata:
-  name: {{ include "llm-d-kserve.fullname" . }}
+  annotations:
+    {{- if .Values.hardwareProfile.create }}
+    opendatahub.io/hardware-profile-name: {{ include "llm-d-kserve.fullname" . | quote }}
+    opendatahub.io/hardware-profile-namespace: {{ .Release.Namespace | quote }}
+    {{- else }}
+    opendatahub.io/hardware-profile-name: {{ .Values.hardwareProfile.name | quote }}
+    opendatahub.io/hardware-profile-namespace: {{ .Values.hardwareProfile.namespace | quote }}
+    {{- end }}
+    opendatahub.io/model-type: generative
+    openshift.io/display-name: {{ .Values.model.name | quote }}
+    security.opendatahub.io/enable-auth: "false"
   labels:
-    {{- include "llm-d-kserve.labels" . | nindent 4 }}
+    opendatahub.io/dashboard: "true"
+  name: {{ include "llm-d-kserve.fullname" . }}
 spec:
   model:
-    name: {{ .Values.model.nameOverride | default (include "llm-d-kserve.fullname" .) }}
-    uri: {{ .Values.model.uri }}
-  replicas: {{ .Values.replicaCount }}
+    name: {{ .Values.model.name | quote }}
+    uri: {{ .Values.model.uri | quote }}
   router:
-    scheduler:
-      template:
-        containers:
-          - name: main
-            args:
-              - -v=4
-              - --poolName
-              - '{{ ChildName .ObjectMeta.Name `-inference-pool` }}'
-              - --poolNamespace
-              - '{{ .ObjectMeta.Namespace }}'
-              - --zap-encoder
-              - json
-              - --grpcPort
-              - "9002"
-              - --grpcHealthPort
-              - "9003"
-              - --secureServing
-              - --modelServerMetricsScheme
-              - https
-              - --modelServerMetricsHttpsInsecureSkipVerify
-              - --certPath
-              - /etc/ssl/certs
-              - --configText
-              - |2
-
-                apiVersion: inference.networking.x-k8s.io/v1alpha1
-                kind: EndpointPickerConfig
-                plugins:
-                  - type: single-profile-handler
-                  - type: max-score-picker
-                  - type: queue-scorer
-                  - type: kv-cache-scorer
-                  - type: prefix-cache-scorer
-                    parameters:
-                      hashBlockSize: 64             # block size 16 * 4 chars / token
-                      maxPrefixBlocksToMatch: 256   # this matches up to ~4k tokens
-                      lruCapacityPerServer: 39744   # total blocks (635904) / block size (16)
-                schedulingProfiles:
-                  - name: default
-                    plugins:
-                      - pluginRef: prefix-cache-scorer
-                        weight: 2.0
-                      - pluginRef: queue-scorer
-                        weight: 1.0
-                      - pluginRef: kv-cache-scorer
-                        weight: 1.0
-                      - pluginRef: max-score-picker
-            
\ No newline at end of file
+    gateway: {}
+    route: {}
+    scheduler: {}
+  replicas: {{ .Values.vllm.replicaCount }}
+  template:
+    {{- if .Values.imagePullSecrets }}
+    imagePullSecrets:
+      {{- toYaml .Values.imagePullSecrets | nindent 6 }}
+    {{- end }}
+    containers:
+    - name: main
+      env:
+        - name: VLLM_ADDITIONAL_ARGS
+          value: {{ .Values.vllm.additionalArguments | quote }}
+      {{- if .Values.vllm.image.repository }}
+      image: {{ include "llm-d-kserve.image" . }}
+      {{- end }}
+      imagePullPolicy: {{ .Values.vllm.image.pullPolicy | quote }}
+      {{- if .Values.vllm.resources }}
+      resources:
+        {{- toYaml .Values.vllm.resources | nindent 8 }}
+      {{- end }}
+    {{- if .Values.vllm.tolerations }}
+    tolerations:
+      {{- toYaml .Values.vllm.tolerations | nindent 6 }}
+    {{- end }}
diff --git a/charts/llm-d-kserve/values.yaml b/charts/llm-d-kserve/values.yaml
index 91faf6a..ca43bfb 100644
--- a/charts/llm-d-kserve/values.yaml
+++ b/charts/llm-d-kserve/values.yaml
@@ -1,170 +1,113 @@
-# Default values for llm-d-kserve.
-# This is a YAML-formatted file.
-# Declare variables to be passed into your templates.
-
-# This will set the replicaset count more information can be found here: https://kubernetes.io/docs/concepts/workloads/controllers/replicaset/
-replicaCount: 1
+# -- String to partially override fullname template (will maintain the release name)
+nameOverride: ""
+# -- String to fully override fullname template
+fullnameOverride: ""
 
 model:
-  nameOverride: "meta-llama/llama-3.2-1b-instruct"
+  # -- The name of the model to be advertised  in the /v1/models endpoint
+  name: "meta-llama/llama-3.2-1b-instruct"
+  # -- The URI of the model to be used by the model server
   uri: oci://quay.io/redhat-ai-services/modelcar-catalog:llama-3.2-1b-instruct
 
-tolerations:
-  - key: "nvidia.com/gpu"
-    operator: "Exists"
-    effect: "NoSchedule"
-
-# This sets the container image more information can be found here: https://kubernetes.io/docs/concepts/containers/images/
-image:
-  repository: nginx
-  # This sets the pull policy for images.
-  pullPolicy: IfNotPresent
-  # Overrides the image tag whose default is the chart appVersion.
-  tag: ""
-
-# This is for the secrets for pulling an image from a private repository more information can be found here: https://kubernetes.io/docs/tasks/configure-pod-container/pull-image-private-registry/
+hardwareProfile:
+  # -- Creates a hardware profile and adds it to the release namespace.
+  create: false
+  # -- The name of the hardware profile to use.
+  # If create is set to true, this value will be ignored and the hardware profile will be created with the release name.
+  name: nvidia-gpu
+  # -- The namespace of the hardware profile to use.
+  # If create is set to true, this value will be ignored and the hardware profile will be created in the release namespace.
+  namespace: redhat-ods-applications
+  # -- The identifiers to add to the hardware profile.
+  # This value will only be used if create is set to true.
+  identifiers: []
+    # - identifier: cpu
+    #   displayName: CPU
+    #   resourceType: CPU
+    #   defaultCount: "4"
+    #   minCount: "1"
+    #   maxCount: "32"
+    # - identifier: memory
+    #   displayName: Memory
+    #   resourceType: Memory
+    #   defaultCount: 8Gi
+    #   minCount: 2Gi
+    #   maxCount: 120Gi
+    # - identifier: nvidia.com/gpu
+    #   displayName: nvidia.com/gpu
+    #   defaultCount: 1
+    #   minCount: 1
+    #   maxCount: 8
+  # -- The tolerations to apply to the hardware profile.
+  # This value will only be used if create is set to true.
+  tolerations: []
+
+
+vllm:
+  # -- The number of replicas to run for the model server
+  replicaCount: 1
+  # -- The tolerations to apply to the model server pod.
+  # If not set, the hardware profile will add the default tolerations.
+  tolerations: []
+    # - key: "nvidia.com/gpu"
+    #   operator: "Exists"
+    #   effect: "NoSchedule"
+  # -- The resources to apply to the model server pod.
+  # If not set, the hardware profile will add the default resources.
+  resources: {}
+    # requests:
+    #   cpu: 1
+    #   memory: 4Gi
+    #   nvidia.com/gpu: 1
+    # limits:
+    #   cpu: 2
+    #   memory: 8Gi
+    #   nvidia.com/gpu: 1
+  # -- The additional arguments to pass to the model server.
+  additionalArguments: ""
+  image:
+    # -- The repository of the image to use for the model server
+    # If not set, the default image will be used.
+    repository: ""
+    # -- The tag of the image to use for the model server
+    tag: ""
+    # -- The pull policy of the image to use for the model server
+    pullPolicy: IfNotPresent
+
+scheduler:
+  endpointPickerConfig: "" #|4
+    # apiVersion: inference.networking.x-k8s.io/v1alpha1
+    # kind: EndpointPickerConfig
+    # plugins:
+    #   - type: pd-profile-handler
+    #     parameters:
+    #       threshold: 30000
+    #   - type: prefill-header-handler
+    #   - type: prefill-filter
+    #   - type: decode-filter
+    #   - type: random-picker
+    #     parameters:
+    #       maxNumOfEndpoints: 1
+    # schedulingProfiles:
+    #   - name: prefill
+    #     plugins:
+    #       - pluginRef: prefill-filter
+    #       - pluginRef: random-picker
+    #   - name: decode
+    #     plugins:
+    #       - pluginRef: decode-filter
+    #       - pluginRef: random-picker
+
+# -- The image pull secrets to use for the model server.
 imagePullSecrets: []
-# This is to override the chart name.
-nameOverride: ""
-fullnameOverride: ""
 
-# This section builds out the service account more information can be found here: https://kubernetes.io/docs/concepts/security/service-accounts/
 serviceAccount:
-  # Specifies whether a service account should be created
+  # -- Specifies whether a service account should be created
   create: true
-  # Automatically mount a ServiceAccount's API credentials?
+  # -- Automatically mount a ServiceAccount's API credentials?
   automount: true
-  # Annotations to add to the service account
+  # -- Annotations to add to the service account
   annotations: {}
-  # The name of the service account to use.
+  # -- The name of the service account to use.
   # If not set and create is true, a name is generated using the fullname template
   name: ""
-
-# This is for setting Kubernetes Annotations to a Pod.
-# For more information checkout: https://kubernetes.io/docs/concepts/overview/working-with-objects/annotations/
-podAnnotations: {}
-# This is for setting Kubernetes Labels to a Pod.
-# For more information checkout: https://kubernetes.io/docs/concepts/overview/working-with-objects/labels/
-podLabels: {}
-
-podSecurityContext: {}
-  # fsGroup: 2000
-
-securityContext: {}
-  # capabilities:
-  #   drop:
-  #   - ALL
-  # readOnlyRootFilesystem: true
-  # runAsNonRoot: true
-  # runAsUser: 1000
-
-# This is for setting up a service more information can be found here: https://kubernetes.io/docs/concepts/services-networking/service/
-service:
-  # This sets the service type more information can be found here: https://kubernetes.io/docs/concepts/services-networking/service/#publishing-services-service-types
-  type: ClusterIP
-  # This sets the ports more information can be found here: https://kubernetes.io/docs/concepts/services-networking/service/#field-spec-ports
-  port: 80
-
-# This block is for setting up the ingress for more information can be found here: https://kubernetes.io/docs/concepts/services-networking/ingress/
-ingress:
-  enabled: false
-  className: ""
-  annotations: {}
-    # kubernetes.io/ingress.class: nginx
-    # kubernetes.io/tls-acme: "true"
-  hosts:
-    - host: chart-example.local
-      paths:
-        - path: /
-          pathType: ImplementationSpecific
-  tls: []
-  #  - secretName: chart-example-tls
-  #    hosts:
-  #      - chart-example.local
-
-# -- Expose the service via gateway-api HTTPRoute
-# Requires Gateway API resources and suitable controller installed within the cluster
-# (see: https://gateway-api.sigs.k8s.io/guides/)
-httpRoute:
-  # HTTPRoute enabled.
-  enabled: false
-  # HTTPRoute annotations.
-  annotations: {}
-  # Which Gateways this Route is attached to.
-  parentRefs:
-  - name: gateway
-    sectionName: http
-    # namespace: default
-  # Hostnames matching HTTP header.
-  hostnames:
-  - chart-example.local
-  # List of rules and filters applied.
-  rules:
-  - matches:
-    - path:
-        type: PathPrefix
-        value: /headers
-  #   filters:
-  #   - type: RequestHeaderModifier
-  #     requestHeaderModifier:
-  #       set:
-  #       - name: My-Overwrite-Header
-  #         value: this-is-the-only-value
-  #       remove:
-  #       - User-Agent
-  # - matches:
-  #   - path:
-  #       type: PathPrefix
-  #       value: /echo
-  #     headers:
-  #     - name: version
-  #       value: v2
-
-resources: {}
-  # We usually recommend not to specify default resources and to leave this as a conscious
-  # choice for the user. This also increases chances charts run on environments with little
-  # resources, such as Minikube. If you do want to specify resources, uncomment the following
-  # lines, adjust them as necessary, and remove the curly braces after 'resources:'.
-  # limits:
-  #   cpu: 100m
-  #   memory: 128Mi
-  # requests:
-  #   cpu: 100m
-  #   memory: 128Mi
-
-# This is to setup the liveness and readiness probes more information can be found here: https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/
-livenessProbe:
-  httpGet:
-    path: /
-    port: http
-readinessProbe:
-  httpGet:
-    path: /
-    port: http
-
-# This section is for setting up autoscaling more information can be found here: https://kubernetes.io/docs/concepts/workloads/autoscaling/
-autoscaling:
-  enabled: false
-  minReplicas: 1
-  maxReplicas: 100
-  targetCPUUtilizationPercentage: 80
-  # targetMemoryUtilizationPercentage: 80
-
-# Additional volumes on the output Deployment definition.
-volumes: []
-# - name: foo
-#   secret:
-#     secretName: mysecret
-#     optional: false
-
-# Additional volumeMounts on the output Deployment definition.
-volumeMounts: []
-# - name: foo
-#   mountPath: "/etc/foo"
-#   readOnly: true
-
-nodeSelector: {}
-
-tolerations: []
-
-affinity: {}

From 579ded69d2559f9fba33687fa81b334c9a5c12a2 Mon Sep 17 00:00:00 2001
From: JeetRH <slele@redhat.com>
Date: Wed, 18 Mar 2026 15:30:29 -0400
Subject: [PATCH 04/12] Minor changes in HWP and values file to get the chart
 working

---
 .../templates/hardwareprofile.yaml            |  2 ++
 charts/llm-d-kserve/values.yaml               | 19 +++++++++++--------
 2 files changed, 13 insertions(+), 8 deletions(-)

diff --git a/charts/llm-d-kserve/templates/hardwareprofile.yaml b/charts/llm-d-kserve/templates/hardwareprofile.yaml
index ba57b5e..fe34d12 100644
--- a/charts/llm-d-kserve/templates/hardwareprofile.yaml
+++ b/charts/llm-d-kserve/templates/hardwareprofile.yaml
@@ -16,6 +16,8 @@ spec:
     node:
       tolerations:
         {{- toYaml .Values.hardwareProfile.tolerations | nindent 8 }}
+    {{- else }}
+    node: {}
     {{- end }}
     type: Node
 {{- end }}
diff --git a/charts/llm-d-kserve/values.yaml b/charts/llm-d-kserve/values.yaml
index ca43bfb..e1c3f4a 100644
--- a/charts/llm-d-kserve/values.yaml
+++ b/charts/llm-d-kserve/values.yaml
@@ -24,24 +24,27 @@ hardwareProfile:
     # - identifier: cpu
     #   displayName: CPU
     #   resourceType: CPU
-    #   defaultCount: "4"
+    #   defaultCount: "2"
     #   minCount: "1"
-    #   maxCount: "32"
+    #   maxCount: "4"
     # - identifier: memory
     #   displayName: Memory
     #   resourceType: Memory
-    #   defaultCount: 8Gi
+    #   defaultCount: 4Gi
     #   minCount: 2Gi
-    #   maxCount: 120Gi
+    #   maxCount: 8Gi
     # - identifier: nvidia.com/gpu
-    #   displayName: nvidia.com/gpu
+    #   displayName: Nvidia Gpu
     #   defaultCount: 1
     #   minCount: 1
-    #   maxCount: 8
+    #   maxCount: 2
+    #   resourceType: Accelerator
   # -- The tolerations to apply to the hardware profile.
   # This value will only be used if create is set to true.
   tolerations: []
-
+    # - key: nvidia.com/gpu
+    #   operator: Exists
+    #   effect: NoSchedule
 
 vllm:
   # -- The number of replicas to run for the model server
@@ -64,7 +67,7 @@ vllm:
     #   memory: 8Gi
     #   nvidia.com/gpu: 1
   # -- The additional arguments to pass to the model server.
-  additionalArguments: ""
+  additionalArguments: "" # "--max-model-len=819"
   image:
     # -- The repository of the image to use for the model server
     # If not set, the default image will be used.

From 3e517515a9026a0f8165db1c85b5ab82fc3bdc4f Mon Sep 17 00:00:00 2001
From: Trevor Royer <troyer@redhat.com>
Date: Mon, 30 Mar 2026 14:25:10 -0600
Subject: [PATCH 05/12] generate helm-docs

---
 charts/llm-d-kserve/README.md | 77 +++++++++++++++++++++++++++++++++++
 1 file changed, 77 insertions(+)
 create mode 100644 charts/llm-d-kserve/README.md

diff --git a/charts/llm-d-kserve/README.md b/charts/llm-d-kserve/README.md
new file mode 100644
index 0000000..99c8a3f
--- /dev/null
+++ b/charts/llm-d-kserve/README.md
@@ -0,0 +1,77 @@
+# llm-d-kserve
+
+A Helm chart for deploying LLM-D with KServe on OpenShift AI
+
+![Version: 0.1.0](https://img.shields.io/badge/Version-0.1.0-informational?style=flat-square) ![Type: application](https://img.shields.io/badge/Type-application-informational?style=flat-square) ![AppVersion: 3.3.0](https://img.shields.io/badge/AppVersion-3.3.0-informational?style=flat-square)
+
+## Installing the Chart
+
+To access charts from this from the cli repository add it:
+
+```sh
+helm repo add redhat-ai-services https://redhat-ai-services.github.io/helm-charts/
+helm repo update redhat-ai-services
+helm upgrade -i [release-name] redhat-ai-services/llm-d-kserve
+```
+
+To include a chart from this repository in an umbrella chart, include it in your dependencies in your `Chart.yaml` file.
+
+```yaml
+apiVersion: v2
+name: example-chart
+description: A Helm chart for Kubernetes
+type: application
+
+version: 0.1.0
+
+appVersion: "1.16.0"
+
+dependencies:
+  - name: "llm-d-kserve"
+    version: "0.1.0"
+    repository: "https://redhat-ai-services.github.io/helm-charts/"
+```
+
+## Usage
+
+### Quick Start
+
+TODO
+
+## Configuration
+
+For a complete list of all configuration options, see the [Values](#values) section below.
+
+## Source Code
+
+* <https://github.com/redhat-ai-services/helm-charts/tree/main/charts/llm-d-kserve>
+
+## Values
+
+| Key | Type | Default | Description |
+|-----|------|---------|-------------|
+| fullnameOverride | string | `""` | String to fully override fullname template |
+| hardwareProfile.create | bool | `false` | Creates a hardware profile and adds it to the release namespace. |
+| hardwareProfile.identifiers | list | `[]` | The identifiers to add to the hardware profile. This value will only be used if create is set to true. |
+| hardwareProfile.name | string | `"nvidia-gpu"` | The name of the hardware profile to use. If create is set to true, this value will be ignored and the hardware profile will be created with the release name. |
+| hardwareProfile.namespace | string | `"redhat-ods-applications"` | The namespace of the hardware profile to use. If create is set to true, this value will be ignored and the hardware profile will be created in the release namespace. |
+| hardwareProfile.tolerations | list | `[]` | The tolerations to apply to the hardware profile. This value will only be used if create is set to true. |
+| imagePullSecrets | list | `[]` | The image pull secrets to use for the model server. |
+| model.name | string | `"meta-llama/llama-3.2-1b-instruct"` | The name of the model to be advertised  in the /v1/models endpoint |
+| model.uri | string | `"oci://quay.io/redhat-ai-services/modelcar-catalog:llama-3.2-1b-instruct"` | The URI of the model to be used by the model server |
+| nameOverride | string | `""` | String to partially override fullname template (will maintain the release name) |
+| scheduler.endpointPickerConfig | string | `""` |  |
+| serviceAccount.annotations | object | `{}` | Annotations to add to the service account |
+| serviceAccount.automount | bool | `true` | Automatically mount a ServiceAccount's API credentials? |
+| serviceAccount.create | bool | `true` | Specifies whether a service account should be created |
+| serviceAccount.name | string | `""` | The name of the service account to use. If not set and create is true, a name is generated using the fullname template |
+| vllm.additionalArguments | string | `""` | The additional arguments to pass to the model server. |
+| vllm.image.pullPolicy | string | `"IfNotPresent"` | The pull policy of the image to use for the model server |
+| vllm.image.repository | string | `""` | The repository of the image to use for the model server If not set, the default image will be used. |
+| vllm.image.tag | string | `""` | The tag of the image to use for the model server |
+| vllm.replicaCount | int | `1` | The number of replicas to run for the model server |
+| vllm.resources | object | `{}` | The resources to apply to the model server pod. If not set, the hardware profile will add the default resources. |
+| vllm.tolerations | list | `[]` | The tolerations to apply to the model server pod. If not set, the hardware profile will add the default tolerations. |
+
+----------------------------------------------
+Autogenerated from chart metadata using [helm-docs v1.14.2](https://github.com/norwoodj/helm-docs/releases/v1.14.2)

From 0ab2081a607c70db51385413fe4ef82c4060f760 Mon Sep 17 00:00:00 2001
From: Trevor Royer <troyer@redhat.com>
Date: Wed, 1 Apr 2026 10:59:31 -0600
Subject: [PATCH 06/12] update tabbing

---
 .../templates/llminferenceservice.yaml        | 30 +++++++++----------
 1 file changed, 15 insertions(+), 15 deletions(-)

diff --git a/charts/llm-d-kserve/templates/llminferenceservice.yaml b/charts/llm-d-kserve/templates/llminferenceservice.yaml
index bff42be..f0311aa 100644
--- a/charts/llm-d-kserve/templates/llminferenceservice.yaml
+++ b/charts/llm-d-kserve/templates/llminferenceservice.yaml
@@ -30,19 +30,19 @@ spec:
       {{- toYaml .Values.imagePullSecrets | nindent 6 }}
     {{- end }}
     containers:
-    - name: main
-      env:
-        - name: VLLM_ADDITIONAL_ARGS
-          value: {{ .Values.vllm.additionalArguments | quote }}
-      {{- if .Values.vllm.image.repository }}
-      image: {{ include "llm-d-kserve.image" . }}
+      - name: main
+        env:
+          - name: VLLM_ADDITIONAL_ARGS
+            value: {{ .Values.vllm.additionalArguments | quote }}
+        {{- if .Values.vllm.image.repository }}
+        image: {{ include "llm-d-kserve.image" . }}
+        {{- end }}
+        imagePullPolicy: {{ .Values.vllm.image.pullPolicy | quote }}
+        {{- if .Values.vllm.resources }}
+        resources:
+          {{- toYaml .Values.vllm.resources | nindent 10 }}
+        {{- end }}
+      {{- if .Values.vllm.tolerations }}
+      tolerations:
+        {{- toYaml .Values.vllm.tolerations | nindent 8 }}
       {{- end }}
-      imagePullPolicy: {{ .Values.vllm.image.pullPolicy | quote }}
-      {{- if .Values.vllm.resources }}
-      resources:
-        {{- toYaml .Values.vllm.resources | nindent 8 }}
-      {{- end }}
-    {{- if .Values.vllm.tolerations }}
-    tolerations:
-      {{- toYaml .Values.vllm.tolerations | nindent 6 }}
-    {{- end }}

From 1cfb28c7ddcff3f9ea35eb1915e03b23d364ffee Mon Sep 17 00:00:00 2001
From: Trevor Royer <troyer@redhat.com>
Date: Mon, 6 Apr 2026 15:27:56 -0600
Subject: [PATCH 07/12] add default topology to screw across multiple nodes

---
 charts/llm-d-kserve/README.md                    |  1 +
 charts/llm-d-kserve/templates/_helpers.tpl       | 13 +++++++++++++
 .../templates/llminferenceservice.yaml           | 16 ++++++++++++++++
 charts/llm-d-kserve/values.yaml                  |  6 ++++++
 4 files changed, 36 insertions(+)

diff --git a/charts/llm-d-kserve/README.md b/charts/llm-d-kserve/README.md
index 99c8a3f..d55e92c 100644
--- a/charts/llm-d-kserve/README.md
+++ b/charts/llm-d-kserve/README.md
@@ -72,6 +72,7 @@ For a complete list of all configuration options, see the [Values](#values) sect
 | vllm.replicaCount | int | `1` | The number of replicas to run for the model server |
 | vllm.resources | object | `{}` | The resources to apply to the model server pod. If not set, the hardware profile will add the default resources. |
 | vllm.tolerations | list | `[]` | The tolerations to apply to the model server pod. If not set, the hardware profile will add the default tolerations. |
+| vllm.topologySpreadConstraints | list | `[{"maxSkew":1,"topologyKey":"kubernetes.io/hostname","whenUnsatisfiable":"ScheduleAnyway"}]` | Pod topology spread constraints for the model server. If labelSelector is omitted on an item, the chart sets matchLabels app.kubernetes.io/name to llm-d-kserve.fullname. |
 
 ----------------------------------------------
 Autogenerated from chart metadata using [helm-docs v1.14.2](https://github.com/norwoodj/helm-docs/releases/v1.14.2)
diff --git a/charts/llm-d-kserve/templates/_helpers.tpl b/charts/llm-d-kserve/templates/_helpers.tpl
index db28079..b0d4d1e 100644
--- a/charts/llm-d-kserve/templates/_helpers.tpl
+++ b/charts/llm-d-kserve/templates/_helpers.tpl
@@ -61,6 +61,19 @@ Create the name of the service account to use
 {{- end }}
 {{- end }}
 
+{{/*
+Render a map as a compact YAML list item ("- key: …" instead of "-\n  key: …").
+*/}}
+{{- define "llm-d-kserve.toYamlListItem" -}}
+{{- $lines := splitList "\n" (toYaml .) -}}
+- {{ first $lines }}
+{{- range rest $lines }}
+{{- if . }}
+  {{ . }}
+{{- end }}
+{{- end }}
+{{- end -}}
+
 {{- define "llm-d-kserve.image" -}}
 {{- if hasPrefix "sha256:" (toString .Values.vllm.image.tag) }}
 {{- printf "%s@%s" .Values.vllm.image.repository .Values.vllm.image.tag }}
diff --git a/charts/llm-d-kserve/templates/llminferenceservice.yaml b/charts/llm-d-kserve/templates/llminferenceservice.yaml
index f0311aa..4c77d26 100644
--- a/charts/llm-d-kserve/templates/llminferenceservice.yaml
+++ b/charts/llm-d-kserve/templates/llminferenceservice.yaml
@@ -29,6 +29,22 @@ spec:
     imagePullSecrets:
       {{- toYaml .Values.imagePullSecrets | nindent 6 }}
     {{- end }}
+    {{- if .Values.vllm.topologySpreadConstraints }}
+    topologySpreadConstraints:
+    {{- range .Values.vllm.topologySpreadConstraints }}
+      {{- if hasKey . "labelSelector" }}
+      {{- /* Caller supplied labelSelector — render unchanged. */}}
+      {{- include "llm-d-kserve.toYamlListItem" . | nindent 6 }}
+      {{- else }}
+      {{- /* No labelSelector — default to matching this chart's pods. */}}
+      {{- $c := deepCopy . }}
+      {{- $defaultMatchLabels := dict "app.kubernetes.io/name" (include "llm-d-kserve.fullname" $) }}
+      {{- $defaultLabelSelector := dict "matchLabels" $defaultMatchLabels }}
+      {{- $_ := set $c "labelSelector" $defaultLabelSelector }}
+      {{- include "llm-d-kserve.toYamlListItem" $c | nindent 6 }}
+      {{- end }}
+    {{- end }}
+    {{- end }}
     containers:
       - name: main
         env:
diff --git a/charts/llm-d-kserve/values.yaml b/charts/llm-d-kserve/values.yaml
index e1c3f4a..850e33f 100644
--- a/charts/llm-d-kserve/values.yaml
+++ b/charts/llm-d-kserve/values.yaml
@@ -76,6 +76,12 @@ vllm:
     tag: ""
     # -- The pull policy of the image to use for the model server
     pullPolicy: IfNotPresent
+  # -- Pod topology spread constraints for the model server.
+  # If labelSelector is omitted on an item, the chart sets matchLabels app.kubernetes.io/name to llm-d-kserve.fullname.
+  topologySpreadConstraints:
+    - maxSkew: 1
+      topologyKey: kubernetes.io/hostname
+      whenUnsatisfiable: ScheduleAnyway
 
 scheduler:
   endpointPickerConfig: "" #|4

From d1aaae58509a21ef8ed4e2256ddc2e990c971588 Mon Sep 17 00:00:00 2001
From: Trevor Royer <troyer@redhat.com>
Date: Mon, 6 Apr 2026 15:33:44 -0600
Subject: [PATCH 08/12] update args logic

---
 .../llm-d-kserve/templates/llminferenceservice.yaml   |  2 +-
 charts/llm-d-kserve/values.yaml                       | 11 ++++++++---
 2 files changed, 9 insertions(+), 4 deletions(-)

diff --git a/charts/llm-d-kserve/templates/llminferenceservice.yaml b/charts/llm-d-kserve/templates/llminferenceservice.yaml
index 4c77d26..72ce8ad 100644
--- a/charts/llm-d-kserve/templates/llminferenceservice.yaml
+++ b/charts/llm-d-kserve/templates/llminferenceservice.yaml
@@ -49,7 +49,7 @@ spec:
       - name: main
         env:
           - name: VLLM_ADDITIONAL_ARGS
-            value: {{ .Values.vllm.additionalArguments | quote }}
+            value: {{ join " " .Values.vllm.additionalArguments | quote }}
         {{- if .Values.vllm.image.repository }}
         image: {{ include "llm-d-kserve.image" . }}
         {{- end }}
diff --git a/charts/llm-d-kserve/values.yaml b/charts/llm-d-kserve/values.yaml
index 850e33f..a4caf70 100644
--- a/charts/llm-d-kserve/values.yaml
+++ b/charts/llm-d-kserve/values.yaml
@@ -56,7 +56,8 @@ vllm:
     #   operator: "Exists"
     #   effect: "NoSchedule"
   # -- The resources to apply to the model server pod.
-  # If not set, the hardware profile will add the default resources.
+  # If not set, the hardware profile will add the default resources, 
+  # however it is highly recommended to set the resources explicitly.
   resources: {}
     # requests:
     #   cpu: 1
@@ -65,9 +66,13 @@ vllm:
     # limits:
     #   cpu: 2
     #   memory: 8Gi
-    #   nvidia.com/gpu: 1
+    #   nvidia.com/gpu: 1\
+
   # -- The additional arguments to pass to the model server.
-  additionalArguments: "" # "--max-model-len=819"
+  additionalArguments: []
+    # - "--max-model-len=8192"
+    # - "--gpu-memory-utilization=0.90"
+
   image:
     # -- The repository of the image to use for the model server
     # If not set, the default image will be used.

From cd13c385f4fb64e180679eabfc97db2a02a48c52 Mon Sep 17 00:00:00 2001
From: Trevor Royer <troyer@redhat.com>
Date: Mon, 6 Apr 2026 15:37:26 -0600
Subject: [PATCH 09/12] add helm labels

---
 charts/llm-d-kserve/templates/hardwareprofile.yaml     | 2 ++
 charts/llm-d-kserve/templates/llminferenceservice.yaml | 3 ++-
 2 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/charts/llm-d-kserve/templates/hardwareprofile.yaml b/charts/llm-d-kserve/templates/hardwareprofile.yaml
index fe34d12..7f99892 100644
--- a/charts/llm-d-kserve/templates/hardwareprofile.yaml
+++ b/charts/llm-d-kserve/templates/hardwareprofile.yaml
@@ -7,6 +7,8 @@ metadata:
     opendatahub.io/description: ""
     opendatahub.io/disabled: "false"
     opendatahub.io/display-name: {{ include "llm-d-kserve.fullname" . | quote }}
+  labels:
+    {{- include "llm-d-kserve.labels" . | nindent 4 }}
   name: {{ include "llm-d-kserve.fullname" . }}
 spec:
   identifiers:
diff --git a/charts/llm-d-kserve/templates/llminferenceservice.yaml b/charts/llm-d-kserve/templates/llminferenceservice.yaml
index 72ce8ad..3fcaf8f 100644
--- a/charts/llm-d-kserve/templates/llminferenceservice.yaml
+++ b/charts/llm-d-kserve/templates/llminferenceservice.yaml
@@ -11,8 +11,9 @@ metadata:
     {{- end }}
     opendatahub.io/model-type: generative
     openshift.io/display-name: {{ .Values.model.name | quote }}
-    security.opendatahub.io/enable-auth: "false"
+    security.opendatahub.io/enable-auth: {{ .Values.auth.enabled | quote }}
   labels:
+    {{- include "llm-d-kserve.labels" . | nindent 4 }}
     opendatahub.io/dashboard: "true"
   name: {{ include "llm-d-kserve.fullname" . }}
 spec:

From 3dfb54f5eacac02dd38fe0a254419454c6ec6f6d Mon Sep 17 00:00:00 2001
From: Trevor Royer <troyer@redhat.com>
Date: Mon, 6 Apr 2026 15:37:46 -0600
Subject: [PATCH 10/12] add role to enable auth

---
 charts/llm-d-kserve/templates/role.yaml | 17 +++++++++++++++++
 charts/llm-d-kserve/values.yaml         | 14 ++++++++++++++
 2 files changed, 31 insertions(+)
 create mode 100644 charts/llm-d-kserve/templates/role.yaml

diff --git a/charts/llm-d-kserve/templates/role.yaml b/charts/llm-d-kserve/templates/role.yaml
new file mode 100644
index 0000000..5d94d7b
--- /dev/null
+++ b/charts/llm-d-kserve/templates/role.yaml
@@ -0,0 +1,17 @@
+{{- if .Values.auth.enabled }}
+kind: Role
+apiVersion: rbac.authorization.k8s.io/v1
+metadata:
+  name: {{ include "llm-d-kserve.fullname" . }}-view
+  labels:
+    {{- include "llm-d-kserve.labels" $ | nindent 4 }}
+rules:
+  - verbs:
+      - get
+    apiGroups:
+      - serving.kserve.io
+    resources:
+      - inferenceservices
+    resourceNames:
+      - {{ include "llm-d-kserve.inferenceServiceName" . }}
+{{- end }}
diff --git a/charts/llm-d-kserve/values.yaml b/charts/llm-d-kserve/values.yaml
index a4caf70..e60baab 100644
--- a/charts/llm-d-kserve/values.yaml
+++ b/charts/llm-d-kserve/values.yaml
@@ -9,6 +9,20 @@ model:
   # -- The URI of the model to be used by the model server
   uri: oci://quay.io/redhat-ai-services/modelcar-catalog:llama-3.2-1b-instruct
 
+auth:
+  # -- Secures the model endpoint and creates a role to grant permissions to service accounts
+  enabled: false
+  # -- Creates service accounts with permissions to access the secured endpoint
+  serviceAccounts: []
+    # - name: my-service-account
+    # - name: existing-service-account
+    #   create: false
+    # - name: existing-in-another-namespace
+    #   namespace: my-other-namespace
+    #   create: false
+    # - name: no-token
+    #   createLegacyToken: false
+
 hardwareProfile:
   # -- Creates a hardware profile and adds it to the release namespace.
   create: false

From 043a873e22531bc7dcd1eaf099f0d85f533f763c Mon Sep 17 00:00:00 2001
From: Trevor Royer <troyer@redhat.com>
Date: Mon, 6 Apr 2026 15:45:21 -0600
Subject: [PATCH 11/12] add ability to provision auth to service accounts

---
 charts/llm-d-kserve/templates/auth-rbac.yaml  | 44 +++++++++++++++++++
 .../templates/{role.yaml => auth-role.yaml}   |  0
 charts/llm-d-kserve/values.yaml               |  3 ++
 3 files changed, 47 insertions(+)
 create mode 100644 charts/llm-d-kserve/templates/auth-rbac.yaml
 rename charts/llm-d-kserve/templates/{role.yaml => auth-role.yaml} (100%)

diff --git a/charts/llm-d-kserve/templates/auth-rbac.yaml b/charts/llm-d-kserve/templates/auth-rbac.yaml
new file mode 100644
index 0000000..d6a00eb
--- /dev/null
+++ b/charts/llm-d-kserve/templates/auth-rbac.yaml
@@ -0,0 +1,44 @@
+{{- if .Values.auth.enabled }}
+{{- range $serviceAccount := .Values.auth.serviceAccounts }}
+{{- if list nil true | has $serviceAccount.create }}
+---
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: {{ $serviceAccount.name }}
+  namespace: {{ $serviceAccount.namespace | default $.Release.Namespace }}
+  labels:
+    {{- include "llm-d-kserve.labels" $ | nindent 4 }}
+{{ if list nil true | has $serviceAccount.createLegacyToken }}
+secrets:
+  - name: {{ $serviceAccount.name }}
+{{- end }}
+automountServiceAccountToken: true
+{{ if list nil true | has $serviceAccount.createLegacyToken }}
+---
+apiVersion: v1
+kind: Secret
+type: kubernetes.io/service-account-token
+metadata:
+  name: {{ $serviceAccount.name }}
+  annotations:
+    kubernetes.io/service-account.name: {{ $serviceAccount.name }}
+{{- end }}
+{{- end }}
+---
+kind: RoleBinding
+apiVersion: rbac.authorization.k8s.io/v1
+metadata:
+  name: {{ include "llm-d-kserve.fullname" $ }}-{{ $serviceAccount.name }}-view
+  labels:
+    {{- include "llm-d-kserve.labels" $ | nindent 4 }}
+subjects:
+  - kind: ServiceAccount
+    name: {{ $serviceAccount.name }}
+    namespace: {{ $serviceAccount.namespace | default $.Release.Namespace }}
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: Role
+  name: {{ include "llm-d-kserve.fullname" $ }}-view
+{{- end }}
+{{- end }}
diff --git a/charts/llm-d-kserve/templates/role.yaml b/charts/llm-d-kserve/templates/auth-role.yaml
similarity index 100%
rename from charts/llm-d-kserve/templates/role.yaml
rename to charts/llm-d-kserve/templates/auth-role.yaml
diff --git a/charts/llm-d-kserve/values.yaml b/charts/llm-d-kserve/values.yaml
index e60baab..0c0a9e3 100644
--- a/charts/llm-d-kserve/values.yaml
+++ b/charts/llm-d-kserve/values.yaml
@@ -13,6 +13,9 @@ auth:
   # -- Secures the model endpoint and creates a role to grant permissions to service accounts
   enabled: false
   # -- Creates service accounts with permissions to access the secured endpoint
+  # If create not set, assume true; if create is set to false, the service account must already exist.
+  # If createLegacyToken not set, assume true; if createLegacyToken is true, a static token will be created for the service account.
+  # If namespace not set, assume the release namespace.
   serviceAccounts: []
     # - name: my-service-account
     # - name: existing-service-account

From 344f93f303bf041d15518202cd0359a4825f65bc Mon Sep 17 00:00:00 2001
From: Trevor Royer <troyer@redhat.com>
Date: Sat, 25 Apr 2026 16:47:41 -0600
Subject: [PATCH 12/12] remove trailing space

---
 charts/llm-d-kserve/values.yaml | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/charts/llm-d-kserve/values.yaml b/charts/llm-d-kserve/values.yaml
index 0c0a9e3..e5275ed 100644
--- a/charts/llm-d-kserve/values.yaml
+++ b/charts/llm-d-kserve/values.yaml
@@ -73,7 +73,7 @@ vllm:
     #   operator: "Exists"
     #   effect: "NoSchedule"
   # -- The resources to apply to the model server pod.
-  # If not set, the hardware profile will add the default resources, 
+  # If not set, the hardware profile will add the default resources,
   # however it is highly recommended to set the resources explicitly.
   resources: {}
     # requests: