Merge pull request #70 from gianlucam76/sync

Sync

Author: gianlucam76

Makefile

@@ -25,7 +25,7 @@ ARCH ?= amd64
 OS ?= $(shell uname -s | tr A-Z a-z)
 K8S_LATEST_VER ?= $(shell curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)
 export CONTROLLER_IMG ?= $(REGISTRY)/$(IMAGE_NAME)
-TAG ?= master
+TAG ?= main
 
 # Get cluster-api version and build ldflags
 clusterapi := $(shell go list -m sigs.k8s.io/cluster-api)

README.md

@@ -5,7 +5,7 @@
 
 # Sveltos
 
-<img src="https://raw.githubusercontent.com/projectsveltos/sveltos-manager/master/logos/logo.png" width="100">
+<img src="https://raw.githubusercontent.com/projectsveltos/sveltos-manager/master/logos/logo.png" width="200">
 
 ## What it is
 Sveltos is tool for policy driven management of kubernetes resources and helm charts in [ClusterAPI](https://github.com/kubernetes-sigs/cluster-api) powered Kubernetes clusters. Sveltos provides declarative APIs to provision  features like Helm charts, ingress controllers, CNIs, storage classes and other resources in a given set of Kubernetes clusters. Sveltos is a freely available and open source. Sveltos is very lightweight and can be installed onto any Kubernetes clusters in minutes.
@@ -70,26 +70,63 @@ To see the full demo, have a look at this [youtube video](https://youtu.be/Ai5Mr
 3. Snapshotting (see [video](https://youtu.be/ALcp1_Nj9r4))
 4. Rollback (see [video](https://youtu.be/sTo6RcWP1BQ))
 5. Conflict detection
-6. Declaritive API and CLI
+6. Declaritive API and CLI [sveltosctl](https://github.com/projectsveltos/sveltosctl)
+
+## Install Sveltos on your laptop using the make cmd
+Just execute `make create-cluster` 
+The command will:
+- create a [KIND](https://sigs.k8s.io/kind) cluster on your laptop;
+- install ClusterAPI;
+- create a CAPI Cluster with Docker as infrastructure provider;
+- install CRD and the Deployment from this project;
+- create a ClusterProfile instance;
+- modify CAPI Cluster labels so to match ClusterProfile selector.
+
+## Install Sveltos on Docker Desktop
+First, enable Kubernetes in Docker Desktop
+```
+1. from the Docker Dashboard, select the Setting icon, or Preferences icon if you use a macOS.
+2. select Kubernetes from the left sidebar.
+3. next to Enable Kubernetes, select the checkbox.
+4. select Apply & Restart to save the settings and then click Install to confirm.
+```
 
-## Test Sveltos on your laptop
-If you want to test it out, just execute, `make create-cluster` and it will:
-1. create a [KIND](https://sigs.k8s.io/kind) cluster;
-2. install ClusterAPI;
-3. create a CAPI Cluster with Docker as infrastructure provider;
-4. install CRD and the Deployment from this project;
-5. create a ClusterProfile instance;
-6. modify CAPI Cluster labels so to match ClusterProfile selector.
+Second, install ClusterAPI
+```
+https://cluster-api.sigs.k8s.io/user/quick-start.html#install-clusterctl
+```
 
-## Install Sveltos 
-If you want to install it on local or remote Kubernetes cluster.
+Third, initialize the management cluster
+```
+clusterctl init --infrastructure docker
+```
 
-1. install ClusterAPI. [ClusterAPI instruction](https://cluster-api.sigs.k8s.io/user/quick-start.html) can be followed.
+Fourth, install Sveltos applying this manifest YAML
+```
+kubectl create -f  https://raw.githubusercontent.com/projectsveltos/cluster-api-feature-manager/master/manifest/manifest.yaml
+```
 
-2. install the CRD and Deployment in the management cluster applying this YAML file:
+Finally, if you have prometheus operator installed in your management cluster, you can apply following ServiceMonitor
+```
+kubectl create -f  https://raw.githubusercontent.com/projectsveltos/cluster-api-feature-manager/master/manifest/service_monitor.yaml
+```
 
+
+## Install Sveltos on any local or remote Kubernetes cluster.
+
+First, install ClusterAPI
 ```
-kubectl create -f  https://raw.githubusercontent.com/projectsveltos/cluster-api-feature-manager/master/manifest/manifest.yaml
+https://cluster-api.sigs.k8s.io/user/quick-start.html#install-clusterctl
+```
+
+Second, initialize the management cluster
+```
+https://cluster-api.sigs.k8s.io/user/quick-start.html#initialize-the-management-cluster
+```
+
+Third, install Sveltos applying this manifest YAML
+```
+kubectl create -f  https://raw.githubusercontent.com/projectsveltos/cluster-api-feature-manager/main/manifest/manifest.yaml
 ```
 
 # Understanding how to configure and use Sveltos
@@ -388,6 +425,13 @@ Another example of misconfiguration is when two different ClusterProfiles match
 In such a case, only one ClusterProfile will be elected and given permission to manage a specific helm release in a given CAPI cluster. Other ClusterProfiles will report such misconfiguration.
 
 
+## Metrics
+Sveltos exposes following metrics:
+1. projectsveltos_program_resources_time_seconds: time to deploy resources in a CAPI clusters (all ClusterProfiles and all CAPI clusters considered);
+2. projectsveltos_program_charts_time_seconds: time to deploy Helm charts in a CAPI clusters (all ClusterProfiles and all CAPI clusters considered);
+3. *clusterNamespace*_*clusterName*_program_resources_time_seconds: time to deploy resources in a specific CAPI clusters (all ClusterProfiles considered);
+4. *clusterNamespace*_*clusterName*_program_charts_time_seconds: time to deploy Helm charts in a specific CAPI clusters (all ClusterProfiles considered);
+
 ### Uninstall CRDs
 To delete the CRDs from the cluster:
 

api/v1alpha1/clusterprofile_types.go

@@ -64,6 +64,12 @@ func (r PolicyRef) String() string {
 	return r.Kind + "-" + r.Namespace + "-" + r.Name
 }
 
+type DryRunReconciliationError struct{}
+
+func (m *DryRunReconciliationError) Error() string {
+	return "mode is DryRun. Nothing is reconciled"
+}
+
 // SyncMode specifies how features are synced in a workload cluster.
 // +kubebuilder:validation:Enum:=OneTime;Continuous;DryRun
 type SyncMode string

api/v1alpha1/zz_generated.deepcopy.go

@@ -8,5 +8,5 @@ spec:
     spec:
       containers:
       # Change the value of image field below to your controller image URL
-      - image: gianlucam76/sveltos-manager-amd64:master
+      - image: gianlucam76/sveltos-manager-amd64:dev
         name: manager

config/default/manager_image_patch.yaml

@@ -198,7 +198,7 @@ func undeployHelmCharts(ctx context.Context, c client.Client,
 		return nil
 	}
 
-	return fmt.Errorf("mode is DryRun. Nothing is reconciled")
+	return &configv1alpha1.DryRunReconciliationError{}
 }
 
 func helmHash(ctx context.Context, c client.Client, clusterSummaryScope *scope.ClusterSummaryScope,
@@ -321,7 +321,7 @@ func handleCharts(ctx context.Context, clusterSummary *configv1alpha1.ClusterSum
 
 	// In DryRun mode always return an error.
 	if clusterSummary.Spec.ClusterProfileSpec.SyncMode == configv1alpha1.SyncModeDryRun {
-		return fmt.Errorf("mode is DryRun. Nothing is reconciled")
+		return &configv1alpha1.DryRunReconciliationError{}
 	}
 	return nil
 }

controllers/handlers_helm.go

@@ -19,6 +19,7 @@ package controllers_test
 import (
 	"context"
 	"crypto/sha256"
+	"errors"
 	"reflect"
 
 	. "github.com/onsi/ginkgo/v2"
@@ -501,7 +502,10 @@ var _ = Describe("HandlersHelm", func() {
 		// such release has never been installed.
 		err = controllers.HandleCharts(context.TODO(), clusterSummary, testEnv.Client, nil, kubeconfig, klogr.New())
 		Expect(err).ToNot(BeNil())
-		Expect(err.Error()).To(Equal("mode is DryRun. Nothing is reconciled"))
+
+		var druRunError *configv1alpha1.DryRunReconciliationError
+		ok := errors.As(err, &druRunError)
+		Expect(ok).To(BeTrue())
 
 		currentClusterReport := &configv1alpha1.ClusterReport{}
 		Eventually(func() bool {

controllers/handlers_helm_test.go

@@ -101,7 +101,7 @@ func deployResources(ctx context.Context, c client.Client,
 	}
 
 	if clusterSummary.Spec.ClusterProfileSpec.SyncMode == configv1alpha1.SyncModeDryRun {
-		return fmt.Errorf("mode is DryRun. Nothing is reconciled")
+		return &configv1alpha1.DryRunReconciliationError{}
 	}
 	return nil
 }
@@ -156,7 +156,7 @@ func undeployResources(ctx context.Context, c client.Client,
 	}
 
 	if clusterSummary.Spec.ClusterProfileSpec.SyncMode == configv1alpha1.SyncModeDryRun {
-		return fmt.Errorf("mode is DryRun. Nothing is reconciled")
+		return &configv1alpha1.DryRunReconciliationError{}
 	}
 	return nil
 }

controllers/handlers_resources.go

@@ -1274,7 +1274,7 @@ spec:
         - --v=5
         command:
         - /manager
-        image: gianlucam76/sveltos-manager-amd64:master
+        image: gianlucam76/sveltos-manager-amd64:main
         livenessProbe:
           httpGet:
             path: /healthz

manifest/manifest.yaml

@@ -0,0 +1,18 @@
+apiVersion: monitoring.coreos.com/v1
+kind: ServiceMonitor
+metadata:
+  labels:
+    control-plane: controller-manager
+  name: fm-controller-manager-metrics-monitor
+  namespace: projectsveltos
+spec:
+  endpoints:
+  - bearerTokenFile: /var/run/secrets/kubernetes.io/serviceaccount/token
+    path: /metrics
+    port: https
+    scheme: https
+    tlsConfig:
+      insecureSkipVerify: true
+  selector:
+    matchLabels:
+      control-plane: controller-manager

manifest/service_monitor.yaml

@@ -0,0 +1,123 @@
+/*
+Copyright 2022. projectsveltos.io. All rights reserved.
+
+Licensed under the Apache License, Version 2.0 (the "License");
+you may not use this file except in compliance with the License.
+You may obtain a copy of the License at
+
+    http://www.apache.org/licenses/LICENSE-2.0
+
+Unless required by applicable law or agreed to in writing, software
+distributed under the License is distributed on an "AS IS" BASIS,
+WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+See the License for the specific language governing permissions and
+limitations under the License.
+*/
+
+package deployer
+
+import (
+	"errors"
+	"fmt"
+	"strings"
+
+	"github.com/go-logr/logr"
+	"github.com/prometheus/client_golang/prometheus"
+	"sigs.k8s.io/controller-runtime/pkg/metrics"
+
+	"github.com/projectsveltos/cluster-api-feature-manager/pkg/logs"
+)
+
+var (
+	programResourceDurationHistogram = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Namespace: "projectsveltos",
+			Name:      "program_resources_time_seconds",
+			Help:      "Program Resources on a workload cluster duration distribution",
+			Buckets:   []float64{1, 10, 30, 60, 120, 180, 240},
+		},
+	)
+
+	programChartDurationHistogram = prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Namespace: "projectsveltos",
+			Name:      "program_charts_time_seconds",
+			Help:      "Program Helm charts on a workload cluster duration distribution",
+			Buckets:   []float64{1, 10, 30, 60, 120, 180, 240},
+		},
+	)
+)
+
+//nolint:gochecknoinits // forced pattern, can't workaround
+func init() {
+	// Register custom metrics with the global prometheus registry
+	metrics.Registry.MustRegister(programResourceDurationHistogram, programChartDurationHistogram)
+}
+
+func newResourceHistogram(clusterNamespace, clusterName string,
+	logger logr.Logger) prometheus.Histogram {
+
+	clusterInfo := strings.ReplaceAll(fmt.Sprintf("%s_%s", clusterNamespace, clusterName), "-", "_")
+	histogram := prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Namespace: clusterInfo,
+			Name:      "program_resources_time_seconds",
+			Help:      "Program Resources on a workload cluster duration distribution",
+			Buckets:   []float64{1, 10, 30, 60, 120, 180, 240},
+		},
+	)
+
+	err := metrics.Registry.Register(histogram)
+	if err != nil {
+		var registrationError *prometheus.AlreadyRegisteredError
+		ok := errors.As(err, &registrationError)
+		if ok {
+			_, ok = registrationError.ExistingCollector.(prometheus.Histogram)
+			if ok {
+				return registrationError.ExistingCollector.(prometheus.Histogram)
+			}
+			logCollectorError(err, logger)
+			return nil
+		}
+		logCollectorError(err, logger)
+		return nil
+	}
+
+	return histogram
+}
+
+func newChartHistogram(clusterNamespace, clusterName string,
+	logger logr.Logger) prometheus.Histogram {
+
+	clusterInfo := strings.ReplaceAll(fmt.Sprintf("%s_%s", clusterNamespace, clusterName), "-", "_")
+	histogram := prometheus.NewHistogram(
+		prometheus.HistogramOpts{
+			Namespace: clusterInfo,
+			Name:      "program_resources_time_seconds",
+			Help:      "Program Helm Charts on a workload cluster duration distribution",
+			Buckets:   []float64{1, 10, 30, 60, 120, 180, 240},
+		},
+	)
+
+	err := metrics.Registry.Register(histogram)
+	if err != nil {
+		var registrationError *prometheus.AlreadyRegisteredError
+		ok := errors.As(err, &registrationError)
+		if ok {
+			_, ok = registrationError.ExistingCollector.(prometheus.Histogram)
+			if ok {
+				return registrationError.ExistingCollector.(prometheus.Histogram)
+			}
+			logCollectorError(err, logger)
+			return nil
+		}
+		logCollectorError(err, logger)
+		return nil
+	}
+
+	return histogram
+}
+
+func logCollectorError(err error, logger logr.Logger) {
+	logger.V(logs.LogInfo).Info(fmt.Sprint("failed to register collector: %w", err))
+}