feat: CID-70 - Operator install preflight checks

Makefile

@@ -118,7 +118,7 @@ test: manifests generate fmt vet setup-envtest ## Run tests.
 # - CERT_MANAGER_INSTALL_SKIP=true
 .PHONY: test-e2e
 test-e2e: manifests generate fmt vet ## Run the e2e tests. Expected an isolated environment using Kind.
-	go test ./test/e2e/ -v -ginkgo.v
+	go test ./test/e2e/ -v -timeout 30m -ginkgo.v -ginkgo.flake-attempts=1
 
 .PHONY: lint
 lint: golangci-lint ## Run golangci-lint linter

charts/castai-castware-operator/templates/cleanup-job.yaml

@@ -1,3 +1,66 @@
+---
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: {{ include "castware-operator.fullname" . }}-cleanup
+  namespace: {{ .Release.Namespace }}
+  labels:
+    {{- include "castware-operator.labels" . | nindent 4 }}
+  annotations:
+    "helm.sh/hook": pre-delete
+    "helm.sh/hook-weight": "-10"
+    "helm.sh/hook-delete-policy": before-hook-creation,hook-succeeded,hook-failed
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRole
+metadata:
+  name: {{ include "castware-operator.fullname" . }}-cleanup
+  labels:
+    {{- include "castware-operator.labels" . | nindent 4 }}
+  annotations:
+    "helm.sh/hook": pre-delete
+    "helm.sh/hook-weight": "-10"
+    "helm.sh/hook-delete-policy": before-hook-creation,hook-succeeded,hook-failed
+rules:
+  - apiGroups:
+      - castware.cast.ai
+    resources:
+      - clusters
+      - components
+    verbs:
+      - get
+      - list
+      - update
+      - patch
+      - delete
+  - apiGroups:
+      - apiextensions.k8s.io
+    resources:
+      - customresourcedefinitions
+    verbs:
+      - delete
+      - get
+      - list
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRoleBinding
+metadata:
+  name: {{ include "castware-operator.fullname" . }}-cleanup
+  labels:
+    {{- include "castware-operator.labels" . | nindent 4 }}
+  annotations:
+    "helm.sh/hook": pre-delete
+    "helm.sh/hook-weight": "-10"
+    "helm.sh/hook-delete-policy": before-hook-creation,hook-succeeded,hook-failed
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: ClusterRole
+  name: {{ include "castware-operator.fullname" . }}-cleanup
+subjects:
+  - kind: ServiceAccount
+    name: {{ include "castware-operator.fullname" . }}-cleanup
+    namespace: {{ .Release.Namespace }}
+---
 apiVersion: batch/v1
 kind: Job
 metadata:
@@ -22,7 +85,7 @@ spec:
         {{- toYaml . | nindent 8 }}
       {{- end }}
 
-      serviceAccountName: {{ include "castware-operator.fullname" . }}-controller-manager
+      serviceAccountName: {{ include "castware-operator.fullname" . }}-cleanup
       restartPolicy: OnFailure
       securityContext:
         {{- toYaml .Values.podSecurityContext | nindent 8 }}

charts/castai-castware-operator/templates/preflight-install-check.yaml

@@ -0,0 +1,82 @@
+{{- if .Values.preflightInstallCheck.enabled }}
+apiVersion: batch/v1
+kind: Job
+metadata:
+  name: {{ include "castware-operator.fullname" . }}-preflight-install-check
+  namespace: {{ .Release.Namespace }}
+  labels:
+    {{- include "castware-operator.labels" . | nindent 4 }}
+  annotations:
+    "helm.sh/hook": pre-install
+    "helm.sh/hook-weight": "1"
+    "helm.sh/hook-delete-policy": before-hook-creation
+spec:
+  backoffLimit: 0
+  activeDeadlineSeconds: 300
+  ttlSecondsAfterFinished: 300
+  template:
+    metadata:
+      name: {{ include "castware-operator.fullname" . }}-preflight-install-check
+      labels:
+        {{- include "castware-operator.selectorLabels" . | nindent 8 }}
+    spec:
+      {{- with .Values.imagePullSecrets }}
+      imagePullSecrets:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      restartPolicy: Never
+      securityContext:
+        {{- toYaml .Values.podSecurityContext | nindent 8 }}
+      containers:
+        - name: {{ .Chart.Name }}
+          image: "{{ .Values.image.repository }}:{{ .Values.image.tag | default .Chart.AppVersion }}"
+          imagePullPolicy: {{ .Values.image.pullPolicy }}
+          args:
+            - preflight-install-check
+          command:
+            - /manager
+          env:
+            - name: POD_NAME
+              valueFrom:
+                fieldRef:
+                  fieldPath: metadata.name
+            - name: POD_NAMESPACE
+              valueFrom:
+                fieldRef:
+                  fieldPath: metadata.namespace
+            - name: OPERATOR_NAME
+              value: {{ include "castware-operator.fullname" . }}
+            - name: HELM_RELEASE_NAME
+              value: {{ .Release.Name }}
+            - name: LOG_LEVEL
+              value: "info"
+            - name: API_URL
+              value: {{ .Values.defaultCluster.api.apiUrl }}
+            - name: HELM_REPO_URL
+              value: {{ .Values.defaultCluster.helmRepoURL }}
+            - name: API_KEY
+              value: {{ .Values.apiKeySecret.apiKey }}
+          resources:
+            {{- toYaml .Values.resources | nindent 12 }}
+          securityContext:
+            {{- toYaml .Values.securityContext | nindent 12 }}
+          volumeMounts:
+            - mountPath: /.cache/helm
+              name: helm-cache
+      {{- with .Values.nodeSelector }}
+      nodeSelector:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      {{- with .Values.affinity }}
+      affinity:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      {{- with .Values.tolerations }}
+      tolerations:
+        {{- toYaml . | nindent 8 }}
+      {{- end }}
+      volumes:
+        - name: helm-cache
+          emptyDir:
+            sizeLimit: 500Mi
+{{- end }}

charts/castai-castware-operator/values.yaml

@@ -81,6 +81,9 @@ crdUpgrade:
 preflightCheck:
   enabled: true
 
+preflightInstallCheck:
+  enabled: true
+
 # Webhook configuration
 webhook:
   enabled: true

cmd/main.go

@@ -3,21 +3,21 @@ package main
 import (
 	"os"
 
-	"github.com/castai/castware-operator/internal/castai"
-	"github.com/castai/castware-operator/internal/config"
 	"github.com/spf13/cobra"
 
-	// Import all Kubernetes client auth plugins (e.g. Azure, GCP, OIDC, etc.)
-	// to ensure that exec-entrypoint and run can make use of them.
-	_ "k8s.io/client-go/plugin/pkg/client/auth"
-
 	apiextensionsv1 "k8s.io/apiextensions-apiserver/pkg/apis/apiextensions/v1"
 	"k8s.io/apimachinery/pkg/runtime"
 	utilruntime "k8s.io/apimachinery/pkg/util/runtime"
 	clientgoscheme "k8s.io/client-go/kubernetes/scheme"
+
+	// Import all Kubernetes client auth plugins (e.g. Azure, GCP, OIDC, etc.)
+	// to ensure that exec-entrypoint and run can make use of them.
+	_ "k8s.io/client-go/plugin/pkg/client/auth"
 	ctrl "sigs.k8s.io/controller-runtime"
 
 	castwarev1alpha1 "github.com/castai/castware-operator/api/v1alpha1"
+	"github.com/castai/castware-operator/internal/castai"
+	"github.com/castai/castware-operator/internal/config"
 	// +kubebuilder:scaffold:imports
 )
 
@@ -70,7 +70,8 @@ func newRootCmd() *cobra.Command {
 
 	rootCmd.Flags().StringVar(&metricsAddr, "metrics-bind-address", "0", "The address the metrics endpoint binds to. "+
 		"Use :8443 for HTTPS or :8080 for HTTP, or leave as 0 to disable the metrics service.")
-	rootCmd.Flags().StringVar(&probeAddr, "health-probe-bind-address", ":8081", "The address the probe endpoint binds to.")
+	rootCmd.Flags().StringVar(&probeAddr, "health-probe-bind-address", ":8081",
+		"The address the probe endpoint binds to.")
 	rootCmd.Flags().BoolVar(&enableLeaderElection, "leader-elect", false,
 		"Enable leader election for controller manager. "+
 			"Enabling this will ensure there is only one active controller manager.")
@@ -80,7 +81,8 @@ func newRootCmd() *cobra.Command {
 		"The directory that contains the metrics server certificate.")
 	rootCmd.Flags().StringVar(&metricsCertName, "metrics-cert-name", "tls.crt",
 		"The name of the metrics server certificate file.")
-	rootCmd.Flags().StringVar(&metricsCertKey, "metrics-cert-key", "tls.key", "The name of the metrics server key file.")
+	rootCmd.Flags().StringVar(&metricsCertKey, "metrics-cert-key", "tls.key",
+		"The name of the metrics server key file.")
 	rootCmd.Flags().BoolVar(&enableHTTP2, "enable-http2", false,
 		"If set, HTTP/2 will be enabled for the metrics and webhook servers")
 
@@ -92,6 +94,7 @@ func main() {
 	rootCmd.AddCommand(newUpgradeCmd())
 	rootCmd.AddCommand(newCleanupCmd())
 	rootCmd.AddCommand(newPreflightCheckCmd())
+	rootCmd.AddCommand(newPreflightInstallCheckCmd())
 
 	version = config.CastwareOperatorVersion{
 		GitCommit: GitCommit,