[CP 1105][CI] Setup hourly build for operator utils image

Makefile

@@ -31,13 +31,25 @@ KMM_BUILDER_IMG ?= gcr.io/kaniko-project/executor:v1.23.2
 KMM_WEBHOOK_IMG_NAME ?= $(DOCKER_REGISTRY)/kernel-module-management-webhook-server
 KMM_OPERATOR_IMG_NAME ?= $(DOCKER_REGISTRY)/kernel-module-management-operator
 
+# Operand related images
+EXPORTER_IMAGE_TAG ?= latest
+METRICS_EXPORTER_IMG = $(DOCKER_REGISTRY)/device-metrics-exporter:$(EXPORTER_IMAGE_TAG)
+DEVICE_CONFIG_MANAGER_IMAGE_TAG ?= latest
+DEVICE_CONFIG_MANAGER_IMG = $(DOCKER_REGISTRY)/device-config-manager:$(DEVICE_CONFIG_MANAGER_IMAGE_TAG)
+TEST_RUNNER_IMAGE_TAG ?= latest
+TEST_RUNNER_IMG = $(DOCKER_REGISTRY)/test-runner:$(TEST_RUNNER_IMAGE_TAG)
+UTILS_IMAGE_TAG ?= latest
+UTILS_IMAGE_NAME ?= $(IMAGE_NAME)-utils
+UTILS_IMG ?= $(DOCKER_REGISTRY)/$(UTILS_IMAGE_NAME):$(UTILS_IMAGE_TAG)
+
 #######################
 # Helm Charts variables
 YAML_FILES=bundle/manifests/amd-gpu-operator-node-metrics_rbac.authorization.k8s.io_v1_rolebinding.yaml bundle/manifests/amd-gpu-operator.clusterserviceversion.yaml bundle/manifests/amd-gpu-operator-node-labeller_rbac.authorization.k8s.io_v1_clusterrolebinding.yaml bundle/manifests/amd-gpu-operator-node-metrics_monitoring.coreos.com_v1_servicemonitor.yaml config/samples/amd.com_deviceconfigs.yaml config/manifests/bases/amd-gpu-operator.clusterserviceversion.yaml example/deviceconfig_example.yaml config/default/kustomization.yaml
 CRD_YAML_FILES = deviceconfig-crd.yaml remediationworkflowstatus-crd.yaml
 K8S_KMM_CRD_YAML_FILES=module-crd.yaml nodemodulesconfig-crd.yaml
 OPENSHIFT_KMM_CRD_YAML_FILES=module-crd.yaml nodemodulesconfig-crd.yaml
 OPENSHIFT_CLUSTER_NFD_CRD_YAML_FILES=nodefeature-crd.yaml nodefeaturediscovery-crd.yaml nodefeaturerule-crd.yaml
+DEFAULT_VALUES_FILES=helm-charts-k8s/values.yaml helm-charts-openshift/values.yaml hack/k8s-patch/metadata-patch/values.yaml hack/openshift-patch/metadata-patch/values.yaml
 
 ifdef OPENSHIFT
 $(info selected openshift)
@@ -199,6 +211,13 @@ update-registry:
 	hack/k8s-patch/metadata-patch/values.yaml helm-charts-k8s/values.yaml \
 	hack/openshift-patch/metadata-patch/values.yaml helm-charts-openshift/values.yaml \
 	example/deviceconfig_example.yaml
+	# update operands image tags
+	@for file in $(DEFAULT_VALUES_FILES); do \
+		yq eval -i '.deviceConfig.spec.metricsExporter.image = "$(METRICS_EXPORTER_IMG)"' $$file; \
+		yq eval -i '.deviceConfig.spec.configManager.image = "$(DEVICE_CONFIG_MANAGER_IMG)"' $$file; \
+		yq eval -i '.deviceConfig.spec.testRunner.image = "$(TEST_RUNNER_IMG)"' $$file; \
+		yq eval -i '.deviceConfig.spec.commonConfig.utilsContainer.image = "$(UTILS_IMG)"' $$file; \
+	done
 	sed -i -e 's|tag:.*$$|tag: ${KMM_IMAGE_TAG}|' \
 	-e 's|repository:.*operator.*$$|repository: ${KMM_OPERATOR_IMG_NAME}|' \
 	-e 's|repository:.*webhook.*$$|repository: ${KMM_WEBHOOK_IMG_NAME}|' \
@@ -298,6 +317,18 @@ docker-push: ## Push docker image with the manager.
 docker-save: ## Save the container image with the manager.
 	docker save $(IMG) | gzip > $(DOCKER_CONTAINER_IMG).tar.gz
 
+.PHONY: docker-build-utils
+docker-build-utils: ## Build docker image for utils container.
+	DOCKER_BUILDKIT=1 docker build -t $(UTILS_IMG) --label HOURLY_TAG=$(HOURLY_TAG_LABEL) -f internal/utils_container/Dockerfile .
+
+.PHONY: docker-push-utils
+docker-push-utils: ## Push docker image for utils container.
+	docker push $(UTILS_IMG)
+
+.PHONY: docker-save-utils
+docker-save-utils: ## Save the utils container image as tar.gz.
+	docker save $(UTILS_IMG) | gzip > $(IMAGE_NAME)-utils-$(IMAGE_TAG).tar.gz
+
 .PHONY: docker-build-env
 docker-build-env: ## Build the docker shell container.
 	@echo "Building the Docker environment..."

hack/k8s-patch/metadata-patch/values.yaml

@@ -88,7 +88,7 @@ deviceConfig:
       initContainerImage: busybox:1.36
       utilsContainer:
         # -- gpu operator utility container image
-        image: docker.io/rocm/gpu-operator-utils:v1.4.0
+        image: docker.io/rocm/gpu-operator-utils:latest
         # -- utility container image pull policy
         imagePullPolicy: IfNotPresent
         # -- utility container image pull secret, e.g. {"name": "mySecretName"}
@@ -129,7 +129,7 @@ deviceConfig:
       # -- external port for pulling metrics from outside the cluster for NodePort service, in the range 30000-32767 (assigned automatically by default)
       nodePort: 32500
       # -- metrics exporter image
-      image: rocm/device-metrics-exporter:latest
+      image: docker.io/rocm/device-metrics-exporter:latest
       # -- metrics exporter image pull policy
       imagePullPolicy: "IfNotPresent"
       # -- name of the metrics exporter config map, e.g. {"name": "metricConfigMapName"}
@@ -187,7 +187,7 @@ deviceConfig:
       # -- enable / disable test runner
       enable: false
       # -- test runner image
-      image: rocm/test-runner:latest
+      image: docker.io/rocm/test-runner:latest
       # -- test runner image pull policy
       imagePullPolicy: "IfNotPresent"
       # -- test runner config map, e.g. {"name": "myConfigMap"}
@@ -214,7 +214,7 @@ deviceConfig:
       # -- enable/disable the config manager 
       enable: false
       # -- config manager image
-      image: rocm/device-config-manager:latest
+      image: docker.io/rocm/device-config-manager:latest
       # -- image pull policy for config manager image
       imagePullPolicy: IfNotPresent
       # -- image pull secret for config manager image, e.g. {"name": "myPullSecret"}

hack/openshift-patch/metadata-patch/values.yaml

@@ -75,8 +75,11 @@ metricsExporter:
 deviceConfig:
   spec:
     metricsExporter:
-      image: rocm/device-metrics-exporter:latest
+      image: docker.io/rocm/device-metrics-exporter:latest
     configManager:
-      image: rocm/device-config-manager:latest
+      image: docker.io/rocm/device-config-manager:latest
     testRunner:
-      image: rocm/test-runner:latest
+      image: docker.io/rocm/test-runner:latest
+    commonConfig:
+      utilsContainer:
+        image: docker.io/rocm/gpu-operator-utils:latest

helm-charts-k8s/README.md

@@ -160,13 +160,13 @@ Kubernetes: `>= 1.29.0-0`
 | crds.defaultCR.install | bool | `true` | Deploy default DeviceConfig during helm chart installation |
 | crds.defaultCR.upgrade | bool | `false` | Deploy / Patch default DeviceConfig during helm chart upgrade. Be careful about this option: 1. Your customized change on default DeviceConfig may be overwritten 2. Your existing DeviceConfig may conflict with upgraded default DeviceConfig  |
 | deviceConfig.spec.commonConfig.initContainerImage | string | `"busybox:1.36"` | init container image |
-| deviceConfig.spec.commonConfig.utilsContainer.image | string | `"docker.io/rocm/gpu-operator-utils:v1.4.0"` | gpu operator utility container image |
+| deviceConfig.spec.commonConfig.utilsContainer.image | string | `"docker.io/rocm/gpu-operator-utils:latest"` | gpu operator utility container image |
 | deviceConfig.spec.commonConfig.utilsContainer.imagePullPolicy | string | `"IfNotPresent"` | utility container image pull policy |
 | deviceConfig.spec.commonConfig.utilsContainer.imageRegistrySecret | object | `{}` | utility container image pull secret, e.g. {"name": "mySecretName"} |
 | deviceConfig.spec.configManager.config | object | `{}` | config map for config manager, e.g. {"name": "myConfigMap"} |
 | deviceConfig.spec.configManager.configManagerTolerations | list | `[]` | config manager tolerations |
 | deviceConfig.spec.configManager.enable | bool | `false` | enable/disable the config manager  |
-| deviceConfig.spec.configManager.image | string | `"rocm/device-config-manager:latest"` | config manager image |
+| deviceConfig.spec.configManager.image | string | `"docker.io/rocm/device-config-manager:latest"` | config manager image |
 | deviceConfig.spec.configManager.imagePullPolicy | string | `"IfNotPresent"` | image pull policy for config manager image |
 | deviceConfig.spec.configManager.imageRegistrySecret | object | `{}` | image pull secret for config manager image, e.g. {"name": "myPullSecret"} |
 | deviceConfig.spec.configManager.selector | object | `{}` | node selector for config manager, if not specified it will reuse spec.selector |
@@ -206,7 +206,7 @@ Kubernetes: `>= 1.29.0-0`
 | deviceConfig.spec.driver.version | string | `"30.20.1"` | specify an out-of-tree driver version to install |
 | deviceConfig.spec.metricsExporter.config | object | `{}` | name of the metrics exporter config map, e.g. {"name": "metricConfigMapName"} |
 | deviceConfig.spec.metricsExporter.enable | bool | `true` | enable / disable device metrics exporter |
-| deviceConfig.spec.metricsExporter.image | string | `"rocm/device-metrics-exporter:latest"` | metrics exporter image |
+| deviceConfig.spec.metricsExporter.image | string | `"docker.io/rocm/device-metrics-exporter:latest"` | metrics exporter image |
 | deviceConfig.spec.metricsExporter.imagePullPolicy | string | `"IfNotPresent"` | metrics exporter image pull policy |
 | deviceConfig.spec.metricsExporter.imageRegistrySecret | object | `{}` | metrics exporter image pull secret, e.g. {"name": "pullSecretName"} |
 | deviceConfig.spec.metricsExporter.nodePort | int | `32500` | external port for pulling metrics from outside the cluster for NodePort service, in the range 30000-32767 (assigned automatically by default) |
@@ -238,7 +238,7 @@ Kubernetes: `>= 1.29.0-0`
 | deviceConfig.spec.selector | object | `{"feature.node.kubernetes.io/amd-gpu":"true"}` | Set node selector for the default DeviceConfig |
 | deviceConfig.spec.testRunner.config | object | `{}` | test runner config map, e.g. {"name": "myConfigMap"} |
 | deviceConfig.spec.testRunner.enable | bool | `false` | enable / disable test runner |
-| deviceConfig.spec.testRunner.image | string | `"rocm/test-runner:latest"` | test runner image |
+| deviceConfig.spec.testRunner.image | string | `"docker.io/rocm/test-runner:latest"` | test runner image |
 | deviceConfig.spec.testRunner.imagePullPolicy | string | `"IfNotPresent"` | test runner image pull policy |
 | deviceConfig.spec.testRunner.imageRegistrySecret | object | `{}` | test runner image pull secret |
 | deviceConfig.spec.testRunner.logsLocation.hostPath | string | `"/var/log/amd-test-runner"` | host directory to save test run logs |

helm-charts-k8s/values.yaml

@@ -88,7 +88,7 @@ deviceConfig:
       initContainerImage: busybox:1.36
       utilsContainer:
         # -- gpu operator utility container image
-        image: docker.io/rocm/gpu-operator-utils:v1.4.0
+        image: docker.io/rocm/gpu-operator-utils:latest
         # -- utility container image pull policy
         imagePullPolicy: IfNotPresent
         # -- utility container image pull secret, e.g. {"name": "mySecretName"}
@@ -129,7 +129,7 @@ deviceConfig:
       # -- external port for pulling metrics from outside the cluster for NodePort service, in the range 30000-32767 (assigned automatically by default)
       nodePort: 32500
       # -- metrics exporter image
-      image: rocm/device-metrics-exporter:latest
+      image: docker.io/rocm/device-metrics-exporter:latest
       # -- metrics exporter image pull policy
       imagePullPolicy: "IfNotPresent"
       # -- name of the metrics exporter config map, e.g. {"name": "metricConfigMapName"}
@@ -187,7 +187,7 @@ deviceConfig:
       # -- enable / disable test runner
       enable: false
       # -- test runner image
-      image: rocm/test-runner:latest
+      image: docker.io/rocm/test-runner:latest
       # -- test runner image pull policy
       imagePullPolicy: "IfNotPresent"
       # -- test runner config map, e.g. {"name": "myConfigMap"}
@@ -214,7 +214,7 @@ deviceConfig:
       # -- enable/disable the config manager 
       enable: false
       # -- config manager image
-      image: rocm/device-config-manager:latest
+      image: docker.io/rocm/device-config-manager:latest
       # -- image pull policy for config manager image
       imagePullPolicy: IfNotPresent
       # -- image pull secret for config manager image, e.g. {"name": "myPullSecret"}

helm-charts-openshift/values.yaml

@@ -75,8 +75,11 @@ metricsExporter:
 deviceConfig:
   spec:
     metricsExporter:
-      image: rocm/device-metrics-exporter:latest
+      image: docker.io/rocm/device-metrics-exporter:latest
     configManager:
-      image: rocm/device-config-manager:latest
+      image: docker.io/rocm/device-config-manager:latest
     testRunner:
-      image: rocm/test-runner:latest
+      image: docker.io/rocm/test-runner:latest
+    commonConfig:
+      utilsContainer:
+        image: docker.io/rocm/gpu-operator-utils:latest