Helm Chart for Kube-AI-Scheduler (#8)

* Helm chart

* Copy CRDs mechanism

* Github workflow build helm chart

Author: omer-dayan

.github/workflows/on-pr.yaml

@@ -16,4 +16,15 @@ jobs:
           go-version: '1.23.4'
 
       - name: Run validation
-        run: make validate
+        run: make validate
+
+      - name: Build helm chart
+        run: |
+          VERSION=$(git rev-parse --short HEAD) 
+          helm package ./deployments/kube-ai-scheduler -d ./charts --version 0.0.0-$VERSION
+
+      - name: Upload helm chart as an artifact
+        uses: actions/upload-artifact@v4
+        with:
+          name: helm-chart
+          path: ./charts/*.tgz

.gitignore

@@ -1,3 +1,4 @@
 bin/
 .idea/
+charts/
 cover.out

Makefile

@@ -8,7 +8,7 @@ $(LOCALBIN):
 CONTROLLER_GEN ?= $(LOCALBIN)/controller-gen
 
 .PHONY: validate
-validate: generate manifests clients
+validate: generate manifests clients copy-crds
 	git diff --exit-code
 
 .PHONY: generate
@@ -18,11 +18,21 @@ generate: controller-gen ## Generate code containing DeepCopy, DeepCopyInto, and
 .PHONY: manifests
 manifests: controller-gen ## Generate ClusterRole and CustomResourceDefinition objects.
 	$(CONTROLLER_GEN) crd:allowDangerousTypes=true,generateEmbeddedObjectMeta=true paths="./pkg/apis/..." output:crd:artifacts:config=config/crd/bases
+	@echo "**********************************************************"
+	@echo "*     NOTE!                                              *"
+	@echo "* Whenever you change pkg/apis/ structures, make sure to *"
+	@echo "* copy them to engine-chart repo.                        *"
+	@echo "* run: 'make copy-crds'                                  *"
+	@echo "**********************************************************"
 
 .PHONY: clients
 clients: ## Generate clients.
 	hack/update-client.sh
 
+.PHONY: copy-crds
+copy-crds:
+	cp -r config/crd/bases/* deployments/kube-ai-scheduler/crds/
+
 .PHONY: controller-gen
 controller-gen: $(CONTROLLER_GEN) ## Download controller-gen locally if necessary.
 $(CONTROLLER_GEN): $(LOCALBIN)

deployments/kube-ai-scheduler/.helmignore

@@ -0,0 +1,36 @@
+# 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/
+
+script/
+deploy/
+helm-charts/*
+subcharts/*
+linux-amd64/*
+helm-*.tar.gz
+operator-sdk
+stable-index/*
+.*
+*.iml
+.circleci/
+.github/

deployments/kube-ai-scheduler/Chart.yaml

@@ -0,0 +1,6 @@
+apiVersion: v2
+name: kube-ai-scheduler
+description: Kube AI Scheduler by NVIDIA
+type: application
+version: 0.0.0
+appVersion: "0.0.0"

deployments/kube-ai-scheduler/crds/scheduling.run.ai_bindrequests.yaml

@@ -0,0 +1,95 @@
+---
+apiVersion: apiextensions.k8s.io/v1
+kind: CustomResourceDefinition
+metadata:
+  annotations:
+    controller-gen.kubebuilder.io/version: v0.16.1
+  name: bindrequests.scheduling.run.ai
+spec:
+  group: scheduling.run.ai
+  names:
+    kind: BindRequest
+    listKind: BindRequestList
+    plural: bindrequests
+    singular: bindrequest
+  scope: Namespaced
+  versions:
+  - name: v1alpha2
+    schema:
+      openAPIV3Schema:
+        description: BindRequest is the Schema for the bindrequests API
+        properties:
+          apiVersion:
+            description: |-
+              APIVersion defines the versioned schema of this representation of an object.
+              Servers should convert recognized schemas to the latest internal value, and
+              may reject unrecognized values.
+              More info: https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#resources
+            type: string
+          kind:
+            description: |-
+              Kind is a string value representing the REST resource this object represents.
+              Servers may infer this from the endpoint the client submits requests to.
+              Cannot be updated.
+              In CamelCase.
+              More info: https://git.k8s.io/community/contributors/devel/sig-architecture/api-conventions.md#types-kinds
+            type: string
+          metadata:
+            type: object
+          spec:
+            description: BindRequestSpec defines the desired state of BindRequest
+            properties:
+              backoffLimit:
+                description: BackoffLimit is the number of retries before giving up
+                format: int32
+                type: integer
+              podName:
+                description: PodName is the name of the pod to bind
+                type: string
+              receivedGPU:
+                description: ReceivedGPU is the amount of GPUs that were received
+                properties:
+                  count:
+                    description: Count is the amount of GPUs devices that were received
+                    type: integer
+                  portion:
+                    description: |-
+                      This is the portion size that the pod will receive from each connected gpu device
+                      This is a serialized float that should be written as a decimal point number.
+                    type: string
+                type: object
+              receivedResourceType:
+                description: ReceivedResourceType is the type of the resource that
+                  was received [Regular/Fraction]
+                type: string
+              selectedGPUGroups:
+                description: |-
+                  SelectedGPUGroups is the name of the selected GPU groups for fractional GPU resources.
+                  Only if the RecievedResourceType is "Fraction"
+                items:
+                  type: string
+                type: array
+              selectedNode:
+                description: SelectedNode is the name of the selected node the pod
+                  should be bound to
+                type: string
+            type: object
+          status:
+            description: BindRequestStatus defines the observed state of BindRequest
+            properties:
+              failedAttempts:
+                description: FailedAttempts is the number of failed attempts
+                format: int32
+                type: integer
+              phase:
+                description: Phase is the current phase of the bindrequest. [Pending/Succeeded/Failed]
+                type: string
+              reason:
+                description: Reason is the reason for the current phase
+                type: string
+            type: object
+        type: object
+    served: true
+    storage: true
+    subresources:
+      status: {}