fix: helm e2e tests were broken

lockwobr · lockwobr · commit 17389fc442ea · 2025-04-18T13:18:52.000-07:00
cleaned up a few refs to nvcr.io that were wrong too
diff --git a/.github/workflows/operator-ci.yaml b/.github/workflows/operator-ci.yaml
@@ -54,21 +54,21 @@ env:
 
 # There is a single job in this workflow. It's configured to run on the latest available version of Ubuntu.
 jobs:
-  unit-test:
+  ## move it all down to the tests job, should do all the tests then, and not double install stuff
+  # unit-test:
+  #   runs-on: ubuntu-latest
+  #   steps:
+  #     - uses: actions/checkout@v4
+  #     - name: Setup Go 1.23
+  #       uses: actions/setup-go@v5
+  #       with:
+  #         go-version: 1.23
+  #     - name: Unit tests
+  #       run: |
+  #         cd operator
+  #         make unit-tests
+  tests:
     runs-on: ubuntu-latest
-    steps:
-      - uses: actions/checkout@v4
-      - name: Setup Go 1.23
-        uses: actions/setup-go@v5
-        with:
-          go-version: 1.23
-      - name: Unit tests
-        run: |
-          cd operator
-          make unit-tests
-  k8s-tests:
-    runs-on: ubuntu-latest
-    needs: [unit-test] # Don't run the k8s tests if the unit tests fail
     steps:
       - uses: actions/checkout@v4
         with:
@@ -94,10 +94,10 @@ jobs:
         run: |
           cd operator
           GITHUB_TOKEN=${{ secrets.github_token }} make create-kind-cluster
-          make e2e-tests
+          make test
   build-and-push-operator:
     runs-on: ubuntu-latest
-    needs: [k8s-tests] # Don't run the build and push if the k8s tests fail
+    needs: [tests] # Don't run the build and push if the k8s tests fail
     # Sets the permissions granted to the `GITHUB_TOKEN` for the actions in this job.
     permissions:
       contents: read
diff --git a/demos/interrupt-wait-for-pod/scr.yaml b/demos/interrupt-wait-for-pod/scr.yaml
@@ -36,8 +36,8 @@ spec:
       app: skyhook-demo-workload
   packages:
     baz:
-      version: 2024.8.10
-      image: nvcr.io/nvidian/swgpu-baseos/shellscript:1.3.1
+      version: 1.1.0
+      image: ghcr.io/nvidia/skyhook-packages/shellscript
       interrupt:
         type: reboot
         # type: service
diff --git a/k8s-tests/chainsaw/helm/helm-chart-test/assert-no-schedule.yaml b/k8s-tests/chainsaw/helm/helm-chart-test/assert-no-schedule.yaml
@@ -53,8 +53,7 @@ spec:
   - command:
     - /manager
     ((env[?name == 'RUNTIME_REQUIRED_TAINT'].value)[0] == 'skyhook.nvidia.com=runtime-required:NoSchedule'): true 
-    image: nvcr.io/nvidian/swgpu-baseos/skyhook-operator:test
-    imagePullPolicy: IfNotPresent
+    image: ghcr.io/nvidia/skyhook/operator:latest ## THIS should change to be like a tag so it can point at a specific commit
     livenessProbe:
       failureThreshold: 3
       httpGet:
@@ -103,7 +102,6 @@ spec:
     - name: KUBERNETES_CLUSTER_DOMAIN
       value: cluster.local
     image: quay.io/brancz/kube-rbac-proxy:v0.15.0
-    imagePullPolicy: IfNotPresent
     name: kube-rbac-proxy
     ports:
     - containerPort: 8443
diff --git a/k8s-tests/chainsaw/helm/helm-chart-test/assert-scheduled.yaml b/k8s-tests/chainsaw/helm/helm-chart-test/assert-scheduled.yaml
@@ -53,7 +53,6 @@ spec:
   - command:
     - /manager
     ((env[?name == 'RUNTIME_REQUIRED_TAINT'].value)[0] == 'skyhook.nvidia.com=runtime-required:NoSchedule'): true
-    imagePullPolicy: IfNotPresent
     livenessProbe:
       failureThreshold: 3
       httpGet:
@@ -102,7 +101,6 @@ spec:
     - name: KUBERNETES_CLUSTER_DOMAIN
       value: cluster.local
     image: quay.io/brancz/kube-rbac-proxy:v0.15.0
-    imagePullPolicy: IfNotPresent
     name: kube-rbac-proxy
     ports:
     - containerPort: 8443
diff --git a/k8s-tests/chainsaw/helm/helm-chart-test/values.yaml b/k8s-tests/chainsaw/helm/helm-chart-test/values.yaml
@@ -29,5 +29,5 @@ controllerManager:
     # telling the helm chart to use the test operator image
     # for more info refer to the README
     image:
-      repository: nvcr.io/nvidian/swgpu-baseos/skyhook-operator
-      tag: test
+      repository: ghcr.io/nvidia/skyhook/operator
+      tag: latest ## THIS should change to be like a tag so it can point at a specific commit
diff --git a/k8s-tests/chainsaw/helm/helm-scale-test/chainsaw-test.yaml b/k8s-tests/chainsaw/helm/helm-scale-test/chainsaw-test.yaml
@@ -22,7 +22,7 @@
 apiVersion: chainsaw.kyverno.io/v1alpha1
 kind: Test
 metadata:
-  name: helm-chart
+  name: helm-scale
 spec:
   description: This test asserts that the helm chart is working as expected. Specifically it asserts that the helm chart works when given a different
     deployment name than skyhook-operator and that the tolerations that are given to the chart through a values file work as expected.
diff --git a/k8s-tests/chainsaw/helm/helm-scale-test/values-scale.yaml b/k8s-tests/chainsaw/helm/helm-scale-test/values-scale.yaml
@@ -23,7 +23,7 @@ controllerManager:
     # telling the helm chart to use the test operator image
     # for more info refer to the README
     image:
-      repository: nvcr.io/nvidian/swgpu-baseos/skyhook-operator
-      tag: test
+      repository: ghcr.io/nvidia/skyhook/operator
+      tag: latest ## THIS should change to be like a tag so it can point at a specific commit
 estimatedNodeCount: 400
 estimatedPackageCount: 5
diff --git a/k8s-tests/chainsaw/helm/helm-scale-test/vaules-override-resources.yaml b/k8s-tests/chainsaw/helm/helm-scale-test/vaules-override-resources.yaml
@@ -30,5 +30,5 @@ controllerManager:
     # telling the helm chart to use the test operator image
     # for more info refer to the README
     image:
-      repository: nvcr.io/nvidian/swgpu-baseos/skyhook-operator
+      repository: ghcr.io/nvidia/skyhook/operator
       tag: test
diff --git a/k8s-tests/chainsaw/helm/install-cert-manager.sh b/k8s-tests/chainsaw/helm/install-cert-manager.sh
@@ -29,7 +29,7 @@ else
     HELM=$(which helm)
 fi
 
-VERSION=${1:-v1.16.2}
+VERSION=${1:-v1.17.0}
 
 ## add chart repo
 $HELM repo add jetstack https://charts.jetstack.io --force-update
diff --git a/k8s-tests/chainsaw/helm/readme.md b/k8s-tests/chainsaw/helm/readme.md
@@ -2,6 +2,6 @@
 This directory holds all the tests for the skyhook operator's helm chart. Right now this mainly ensures that tolerations set in the helm chart actually work and that the operator can be deployed successfully under another deployment name than skyhook-operator.
 
 ## Test Image
-The image that is used by these tests should be `nvcr.io/nvidian/swgpu-baseos/skyhook-operator:test` since this will be built in CI every time a commit is pushed to Gitlab and will make sure that you current changes to the operator are compatible with the helm chart still.
+The image that is used by these tests should be ` ghcr.io/nvidia/skyhook/operator:test` (NOTE: this does not exist in the current github CI, this needs to be fixed) since this will be built in CI every time a commit is pushed to Gitlab and will make sure that you current changes to the operator are compatible with the helm chart still. 
 
 **NOTE:** When you run the helm chart tests locally it may be using an outdated version of the test image since it hasn't been pushed and built by the CI. Be careful in the assumptions you make as your changes to the operator may pass the helm chart tests locally but fail in CI.
diff --git a/operator/Makefile b/operator/Makefile
@@ -19,7 +19,7 @@ VERSION ?= $(GIT_TAG_LAST)
 
 # Image URL to use all building/pushing image targets
 ## TODO: update this to the correct image location
-IMG ?= nvcr.io/nvidian/swgpu-baseos/skyhook-operator:latest
+IMG ?= ghcr.io/nvidia/skyhook/operator:latest
 
 ## default version of kind to use
 KIND_VERSION?=1.30.10
@@ -101,6 +101,7 @@ confirm:
 .PHONY: manifests
 manifests: controller-gen ## Generate WebhookConfiguration, ClusterRole and CustomResourceDefinition objects.
 	$(CONTROLLER_GEN) rbac:roleName=manager-role crd webhook paths="./..." output:crd:artifacts:config=config/crd/bases
+	$(MAKE) license-fmt ## fix up license headers
 
 .PHONY: generate
 generate: controller-gen ## Generate code containing DeepCopy, DeepCopyInto, and DeepCopyObject method implementations.
diff --git a/operator/api/v1alpha1/zz_generated.deepcopy.go b/operator/api/v1alpha1/zz_generated.deepcopy.go
diff --git a/operator/config/manager/kustomization.yaml b/operator/config/manager/kustomization.yaml
@@ -24,5 +24,5 @@ apiVersion: kustomize.config.k8s.io/v1beta1
 kind: Kustomization
 images:
 - name: controller
-  newName: nvcr.io/nvidian/swgpu-baseos/skyhook-operator
+  newName: ghcr.io/nvidia/skyhook/operator
   newTag: latest
diff --git a/operator/config/manager/manager.yaml b/operator/config/manager/manager.yaml
@@ -112,7 +112,7 @@ spec:
           - name: PAUSE_IMAGE
             value: registry.k8s.io/pause:3.10
           - name: AGENT_IMAGE
-            value: nvcr.io/nvidian/swgpu-baseos/skyhook-agent:latest
+            value: ghcr.io/nvidia/skyhook/agent:latest
         image: controller:latest
         name: manager
         securityContext:
diff --git a/operator/internal/controller/skyhook_controller.go b/operator/internal/controller/skyhook_controller.go
@@ -81,7 +81,7 @@ type SkyhookOperatorOptions struct {
 	ReapplyOnReboot      bool          `env:"REAPPLY_ON_REBOOT, default=false"`
 	RuntimeRequiredTaint string        `env:"RUNTIME_REQUIRED_TAINT, default=skyhook.nvidia.com=runtime-required:NoSchedule"`
 	PauseImage           string        `env:"PAUSE_IMAGE, default=registry.k8s.io/pause:3.10"`
-	AgentImage           string        `env:"AGENT_IMAGE, default=nvcr.io/nvidian/swgpu-baseos/skyhook-agent:latest"` // TODO: this needs to be updated with a working default
+	AgentImage           string        `env:"AGENT_IMAGE, default=ghcr.io/nvidia/skyhook/agent:latest"` // TODO: this needs to be updated with a working default
 }
 
 func (o *SkyhookOperatorOptions) Validate() error {
