Add reconciliation status + always update status of managed resources. (#76)

Author: Gerrit91

.github/workflows/docker.yaml

@@ -38,7 +38,7 @@ jobs:
         cache: false
 
     - name: Lint
-      uses: golangci/golangci-lint-action@v4
+      uses: golangci/golangci-lint-action@v5
       with:
         args: --build-tags integration -p bugs -p unused --timeout=5m
 
@@ -48,7 +48,12 @@ jobs:
         [ "${GITHUB_EVENT_NAME}" == 'release' ] && echo "tag=${GITHUB_REF##*/}" >> $GITHUB_ENV || true
         [ "${GITHUB_EVENT_NAME}" == 'push' ] && echo "tag=latest" >> $GITHUB_ENV || true
 
-    - name: Build and push image
+    - name: Test and build
+      run: |
+        make test
+        make
+
+    - name: Push image
       uses: docker/build-push-action@v5
       with:
         context: .

.gitignore

@@ -5,3 +5,5 @@ api/durosv2/*
 bin/
 **/*/statik.go
 .vscode
+# Output of the go coverage tool, specifically when used with LiteIDE
+*.out

Dockerfile

@@ -1,36 +1,5 @@
-# Build the duros-controller binary
-FROM golang:1.22 as builder
-
-ENV KUBEBUILDER_DOWNLOAD_URL=https://github.com/kubernetes-sigs/kubebuilder/releases/download
-ENV KUBEBUILDER_VER=v3.14.0
-RUN set -ex \
- && mkdir -p /usr/local/bin \
- && curl -L ${KUBEBUILDER_DOWNLOAD_URL}/v${KUBEBUILDER_VER}/kubebuilder_linux_amd64 -o /usr/local/bin/kubebuilder \
- && chmod +x /usr/local/bin/kubebuilder
-
-WORKDIR /workspace
-# Copy the Go Modules manifests
-COPY go.mod go.mod
-COPY go.sum go.sum
-# cache deps before building and copying source so that we don't need to re-download as much
-# and so that source changes don't invalidate our downloaded layer
-RUN go mod download
-
-# Copy the go source
-COPY .git/ .git/
-COPY Makefile Makefile
-COPY api/ api/
-COPY controllers/ controllers/
-COPY hack/ hack/
-COPY config/ config/
-COPY main.go main.go
-
-# Build
-RUN make
-
-# Final Image
 FROM alpine:3.19
 WORKDIR /
-COPY --from=builder /workspace/bin/duros-controller .
+COPY bin/duros-controller .
 USER 65534
 ENTRYPOINT ["/duros-controller"]

Makefile

@@ -10,23 +10,23 @@ GITVERSION := $(shell git describe --long --all)
 BUILDDATE := $(shell date -Iseconds)
 VERSION := $(or ${VERSION},devel)
 
-# Get the currently used golang install path (in GOPATH/bin, unless GOBIN is set)
-ifeq (,$(shell go env GOBIN))
-GOBIN=$(shell go env GOPATH)/bin
-else
-GOBIN=$(shell go env GOBIN)
-endif
+CONTROLLER_TOOLS_VERSION ?= v0.14.0
+LOCALBIN ?= $(shell pwd)/bin
+CONTROLLER_GEN ?= $(LOCALBIN)/controller-gen
 
 all: manager
 
+$(LOCALBIN):
+	mkdir -p $(LOCALBIN)
+
 # clean generated code
 .PHONY: clean
 clean:
 	rm -f api/v1/*pb.go bin/*
 
 # Run tests
 test: generate fmt vet manifests
-	go test ./... -coverprofile cover.out -race -timeout 30s
+	CGO_ENABLED=1 go test ./... -coverprofile cover.out -race -timeout 30s
 
 # Build manager binary
 manager: generate fmt vet
@@ -78,22 +78,11 @@ docker-build: test
 docker-push:
 	docker push ${IMG}
 
-# find or download controller-gen
-# download controller-gen if necessary
-controller-gen:
-ifeq (, $(shell which controller-gen))
-	@{ \
-	set -e ;\
-	CONTROLLER_GEN_TMP_DIR=$$(mktemp -d) ;\
-	cd $$CONTROLLER_GEN_TMP_DIR ;\
-	go mod init tmp ;\
-	go install sigs.k8s.io/controller-tools/cmd/controller-gen@v0.14.0 ;\
-	rm -rf $$CONTROLLER_GEN_TMP_DIR ;\
-	}
-CONTROLLER_GEN=$(GOBIN)/controller-gen
-else
-CONTROLLER_GEN=$(shell which controller-gen)
-endif
+.PHONY: controller-gen
+controller-gen: $(CONTROLLER_GEN)
+$(CONTROLLER_GEN): $(LOCALBIN)
+	test -s $(LOCALBIN)/controller-gen && $(LOCALBIN)/controller-gen --version | grep -q $(CONTROLLER_TOOLS_VERSION) || \
+	GOBIN=$(LOCALBIN) go install sigs.k8s.io/controller-tools/cmd/controller-gen@$(CONTROLLER_TOOLS_VERSION)
 
 .PHONY: certs
 certs:

api/v1/duros_types.go

@@ -55,6 +55,8 @@ type DurosSpec struct {
 
 // DurosStatus defines the observed state of Duros
 type DurosStatus struct {
+	// ReconcileStatus describes the current status of the reconciliation
+	ReconcileStatus ReconcileStatus `json:"reconcileStatus" description:"The current status of the reconciliation of this resource"`
 	// ManagedResourceStatuses contains a list of statuses of resources managed by this controller
 	ManagedResourceStatuses []ManagedResourceStatus `json:"managedResourceStatuses" description:"A list of managed resource statuses"`
 }
@@ -72,6 +74,13 @@ type ManagedResourceStatus struct {
 	LastUpdateTime metav1.Time `json:"lastUpdateTime" description:"The time when this status was last updated"`
 }
 
+type ReconcileStatus struct {
+	// LastReconcile is the point in time when the last reconciliation took place
+	LastReconcile *metav1.Time `json:"lastReconcile,omitempty" description:"Timestamp of last reconciliation"`
+	// Error indicates a reconciliation error that occurred, it's empty when there is no error
+	Error *string `json:"error,omitempty" description:"If set describes that there is an error during reconciliation"`
+}
+
 // HealthState describes the state of a managed resource
 type HealthState string
 

api/v1/zz_generated.deepcopy.go

@@ -109,8 +109,22 @@ spec:
                   - state
                   type: object
                 type: array
+              reconcileStatus:
+                description: ReconcileStatus describes the current status of the reconciliation
+                properties:
+                  error:
+                    description: Error indicates a reconciliation error that occurred,
+                      it's empty when there is no error
+                    type: string
+                  lastReconcile:
+                    description: LastReconcile is the point in time when the last
+                      reconciliation took place
+                    format: date-time
+                    type: string
+                type: object
             required:
             - managedResourceStatuses
+            - reconcileStatus
             type: object
         type: object
     served: true

config/crd/bases/storage.metal-stack.io_duros.yaml

@@ -89,10 +89,36 @@ func (r *DurosReconciler) Reconcile(ctx context.Context, req ctrl.Request) (ctrl
 		return ctrl.Result{}, nil
 	}
 
-	err := validateDuros(duros)
+	var err error
+
+	defer func() {
+		now := metav1.NewTime(time.Now())
+
+		duros.Status.ReconcileStatus = duroscontrollerv1.ReconcileStatus{
+			LastReconcile: &now,
+			Error:         nil,
+		}
+
+		if err != nil {
+			msg := err.Error()
+			duros.Status.ReconcileStatus.Error = &msg
+		}
+
+		r.setManagedResourceStatus(ctx, duros)
+
+		if err := r.Status().Update(ctx, duros); err != nil {
+			log.Error(err, "error updating status of duros resource", "name", duros.Name)
+			return
+		}
+
+		r.Log.Info("status updated", "name", duros.Name)
+	}()
+
+	err = validateDuros(duros)
 	if err != nil {
 		return requeue, err
 	}
+
 	projectID := duros.Spec.MetalProjectID
 	storageClasses := duros.Spec.StorageClasses
 
@@ -118,35 +144,38 @@ func (r *DurosReconciler) Reconcile(ctx context.Context, req ctrl.Request) (ctrl
 		return requeue, err
 	}
 
-	err = r.reconcileStatus(ctx, duros)
-	if err != nil {
-		return requeue, err
-	}
-
 	return ctrl.Result{
 		// we requeue in a small interval to ensure resources are recreated quickly
 		// and status is updated regularly
 		RequeueAfter: 30 * time.Second,
 	}, nil
 }
 
-func (r *DurosReconciler) reconcileStatus(ctx context.Context, duros *duroscontrollerv1.Duros) error {
+func (r *DurosReconciler) setManagedResourceStatus(ctx context.Context, duros *duroscontrollerv1.Duros) {
 	var (
 		updateTime = metav1.NewTime(time.Now())
 		ds         = &appsv1.DaemonSet{}
 		sts        = &appsv1.StatefulSet{}
+
+		dsState = duroscontrollerv1.HealthStateRunning
+		dsMsg   = "All replicas are ready"
+
+		stsState = duroscontrollerv1.HealthStateRunning
+		stsMsg   = "All replicas are ready"
 	)
 
 	err := r.Shoot.Get(ctx, types.NamespacedName{Name: lbCSINodeName, Namespace: namespace}, ds)
 	if err != nil {
-		return fmt.Errorf("error getting daemon set: %w", err)
+		r.Log.Error(err, "error getting daemon set")
+		dsState = duroscontrollerv1.HealthStateNotRunning
+		dsMsg = err.Error()
 	}
 
 	dsStatus := duroscontrollerv1.ManagedResourceStatus{
 		Name:           ds.Name,
 		Group:          "DaemonSet", // ds.GetObjectKind().GroupVersionKind().String() --> this does not work :(
-		State:          duroscontrollerv1.HealthStateRunning,
-		Description:    "All replicas are ready",
+		State:          dsState,
+		Description:    dsMsg,
 		LastUpdateTime: updateTime,
 	}
 
@@ -157,14 +186,16 @@ func (r *DurosReconciler) reconcileStatus(ctx context.Context, duros *duroscontr
 
 	err = r.Shoot.Get(ctx, types.NamespacedName{Name: lbCSIControllerName, Namespace: namespace}, sts)
 	if err != nil {
-		return fmt.Errorf("error getting statefulset: %w", err)
+		r.Log.Error(err, "error getting stateful set")
+		stsState = duroscontrollerv1.HealthStateNotRunning
+		stsMsg = err.Error()
 	}
 
 	stsStatus := duroscontrollerv1.ManagedResourceStatus{
 		Name:           sts.Name,
 		Group:          "StatefulSet", // sts.GetObjectKind().GroupVersionKind().String() --> this does not work :(
-		State:          duroscontrollerv1.HealthStateRunning,
-		Description:    "All replicas are ready",
+		State:          stsState,
+		Description:    stsMsg,
 		LastUpdateTime: updateTime,
 	}
 
@@ -179,14 +210,6 @@ func (r *DurosReconciler) reconcileStatus(ctx context.Context, duros *duroscontr
 	}
 
 	duros.Status.ManagedResourceStatuses = []duroscontrollerv1.ManagedResourceStatus{dsStatus, stsStatus}
-	err = r.Status().Update(ctx, duros)
-	if err != nil {
-		return fmt.Errorf("error updating status: %w", err)
-	}
-
-	r.Log.Info("status updated", "name", duros.Name)
-
-	return nil
 }
 
 // SetupWithManager boilerplate to setup the Reconciler