fix: testing fixes

.vscode/settings.json

@@ -0,0 +1,5 @@
+{
+    "githubPullRequests.ignoredPullRequestBranches": [
+        "main"
+    ]
+}

README.md

@@ -74,7 +74,7 @@ We will build and publish our resolver changes.
 1. Go into resolver directory. 
 2. Run build and publish command.
 ```bash
-make docker-build-resolver docker-publish-resolver IMG=ramantehlan/elasti-resolver:latest  
+make docker-buildx-resolver IMG=ramantehlan/elasti-resolver:v1alpha1
 ```
 
 ### Build Operator
@@ -84,14 +84,21 @@ We will build and publish our Operator changes.
 1. Go into operator directory.
 2. Run build and publish command.
 ```bash
-make docker-build-resolver docker-publish-resolver IMG=ramantehlan/elasti-resolver:latest
+make docker-buildx IMG=ramantehlan/elasti-operator:v1alpha1
 ```
 
 > Once your changes are published, you can re-deploy in your cluster.
 
 # Configuration
 TBA
 
+# Playground 
+
+```
+ make docker-build docker-publish IMG=localhost:5001/elasti-operator:v1alpha1
+ make docker-build docker-publish IMG=localhost:5001/elasti-resolver:v1alpha1
+```
+
 # Icon 
 
 The icon is <a href="https://www.flaticon.com/free-icons/full-screen" title="full-screen icons">Full-screen icon created by Uniconlabs - Flaticon</a>. 

demo-elastiService.yaml

@@ -0,0 +1,23 @@
+apiVersion: elasti.truefoundry.com/v1alpha1
+kind: ElastiService
+metadata:
+  labels:
+    app.kubernetes.io/name: elasti-operator
+    app.kubernetes.io/managed-by: kustomize
+  name: emotion-class
+  namespace: raman-ws
+spec:
+  queueTimeout: 4
+  idlePeriod: 20
+  service: emotion-class-svc
+  deploymentName: emotion-class-svc
+  minTargetReplicas: 1
+  # scaleTargetRef:
+  #   apiVersion: apps/v1
+  #   kind: Rollouts
+  #   name: emotion-class-svc
+  scaleTargetRef:
+    apiVersion: apps/v1
+    kind: Deployments
+    name: emotion-class-svc
+

install.yaml

@@ -100,6 +100,15 @@ metadata:
   name: elasti-operator-controller-manager
   namespace: elasti-operator-system
 ---
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  labels:
+    app.kubernetes.io/managed-by: kustomize
+    app.kubernetes.io/name: elasti-resolver
+  name: elasti-resolver
+  namespace: elasti
+---
 apiVersion: rbac.authorization.k8s.io/v1
 kind: Role
 metadata:
@@ -156,6 +165,16 @@ rules:
   - watch
   - update
   - patch
+- apiGroups:
+  - apps
+  resources:
+  - rollouts
+  verbs:
+  - get
+  - list
+  - watch
+  - update
+  - patch
 - apiGroups:
   - discovery.k8s.io
   resources:
@@ -297,7 +316,7 @@ roleRef:
   name: elasti-operator-leader-election-role
 subjects:
 - kind: ServiceAccount
-  name: elasti-operator-controller-manager
+  name: elasti-resolver
   namespace: elasti-operator-system
 ---
 apiVersion: rbac.authorization.k8s.io/v1
@@ -318,6 +337,22 @@ subjects:
 ---
 apiVersion: rbac.authorization.k8s.io/v1
 kind: ClusterRoleBinding
+metadata:
+  labels:
+    app.kubernetes.io/managed-by: kustomize
+    app.kubernetes.io/name: elasti-operator
+  name: elasti-resolver-additional-access-binding
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: ClusterRole
+  name: elasti-operator-additional-access
+subjects:
+- kind: ServiceAccount
+  name: elasti-operator-controller-manager
+  namespace: elasti
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: ClusterRoleBinding
 metadata:
   labels:
     app.kubernetes.io/managed-by: kustomize
@@ -484,7 +519,7 @@ spec:
     spec:
       containers:
       - name: playground
-        image: ramantehlan/elasti-resolver:latest
+        image: ramantehlan/elasti-resolver:v1alpha1
         env:
         - name: SYSTEM_NAMESPACE
           value: elasti

operator/Makefile

@@ -103,14 +103,14 @@ docker-push: ## Push docker image with the manager.
 # - have enabled BuildKit. More info: https://docs.docker.com/develop/develop-images/build_enhancements/
 # - be able to push the image to your registry (i.e. if you do not set a valid value via IMG=<myregistry/image:<tag>> then the export will fail)
 # To adequately provide solutions that are compatible with multiple platforms, you should consider using this option.
-PLATFORMS ?= linux/arm64,linux/amd64,linux/s390x,linux/ppc64le
+PLATFORMS ?= linux/arm64,linux/amd64 #,linux/s390x,linux/ppc64le
 .PHONY: docker-buildx
 docker-buildx: ## Build and push docker image for the manager for cross-platform support
 	# copy existing Dockerfile and insert --platform=${BUILDPLATFORM} into Dockerfile.cross, and preserve the original Dockerfile
 	sed -e '1 s/\(^FROM\)/FROM --platform=\$$\{BUILDPLATFORM\}/; t' -e ' 1,// s//FROM --platform=\$$\{BUILDPLATFORM\}/' Dockerfile > Dockerfile.cross
 	- $(CONTAINER_TOOL) buildx create --name project-v3-builder
 	$(CONTAINER_TOOL) buildx use project-v3-builder
-	- $(CONTAINER_TOOL) buildx build --push --platform=$(PLATFORMS) --tag ${IMG} -f Dockerfile.cross .
+	- $(CONTAINER_TOOL) buildx build --push --platform=$(PLATFORMS) --tag ${IMG} -f Dockerfile.cross ../
 	- $(CONTAINER_TOOL) buildx rm project-v3-builder
 	rm Dockerfile.cross
 
@@ -141,6 +141,7 @@ deploy: manifests kustomize ## Deploy controller to the K8s cluster specified in
 
 .PHONY: undeploy
 undeploy: kustomize ## Undeploy controller from the K8s cluster specified in ~/.kube/config. Call with ignore-not-found=true to ignore resource not found errors during deletion.
+	# $(KUBECTL) delete --ignore-not-found=true -f ../playground/config/watch-crd.yaml
 	$(KUSTOMIZE) build config/default | $(KUBECTL) delete --ignore-not-found=$(ignore-not-found) -f -
 
 ##@ Dependencies

operator/README.md

@@ -21,7 +21,7 @@ make docker-build docker-push IMG=<some-registry>/elasti-operator:tag
 
 **NOTE:** This image ought to be published in the personal registry you specified.
 And it is required to have access to pull the image from the working environment.
-Make sure you have the proper permission to the registry if the above commands don’t work.
+Make sure you have the proper permission to the registry if the above commands don't work.
 
 **Install the CRDs into the cluster:**
 

operator/api/v1alpha1/elastiservice_types.go

@@ -31,12 +31,12 @@ const (
 type ElastiServiceSpec struct {
 	// INSERT ADDITIONAL SPEC FIELDS - desired state of cluster
 	// Important: Run "make" to regenerate code after modifying this file
-	ScaleTargetRef     ScaleTargetRef `json:"scaleTargetRef,omitempty"`
-	Service            string         `json:"service,omitempty"`
-	DeploymentName     string         `json:"deploymentName,omitempty"`
-	QTimout            int32          `json:"queueTimeout,omitempty"`
-	IdlePeriod         int32          `json:"idlePeriod,omitempty"`
-	ServeReplicasCount int32          `json:"minReplicas,omitempty"`
+	ScaleTargetRef    ScaleTargetRef `json:"scaleTargetRef,omitempty"`
+	Service           string         `json:"service,omitempty"`
+	DeploymentName    string         `json:"deploymentName,omitempty"`
+	QTimout           int32          `json:"queueTimeout,omitempty"`
+	IdlePeriod        int32          `json:"idlePeriod,omitempty"`
+	MinTargetReplicas int32          `json:"minTargetReplicas,omitempty"`
 }
 
 type ScaleTargetRef struct {

operator/config/crd/bases/elasti.truefoundry.com_elastiservices.yaml

@@ -44,7 +44,7 @@ spec:
               idlePeriod:
                 format: int32
                 type: integer
-              minReplicas:
+              minTargetReplicas:
                 format: int32
                 type: integer
               queueTimeout:

operator/config/manager/kustomization.yaml

@@ -5,5 +5,5 @@ apiVersion: kustomize.config.k8s.io/v1beta1
 kind: Kustomization
 images:
 - name: controller
-  newName: localhost:5001/elasti-operator
-  newTag: v1alpha1
+  newName: controller
+  newTag: latest

operator/config/manager/manager.yaml

@@ -64,7 +64,7 @@ spec:
           - --leader-elect
           - --health-probe-bind-address=:8081
           - --metrics-bind-address=0
-        image: controller:latest
+        image: localhost:5001/elasti-operator:v1alpha1
         name: manager
         ports:
         - containerPort: 8013

operator/internal/controller/elastiservice_controller.go

@@ -24,21 +24,19 @@ import (
 )
 
 type (
-	RunReconcileFunc        func(ctx context.Context, req ctrl.Request, mode string) (res ctrl.Result, err error)
+	SwitchModeFunc          func(ctx context.Context, req ctrl.Request, mode string) (res ctrl.Result, err error)
 	ElastiServiceReconciler struct {
 		client.Client
-		Scheme            *kRuntime.Scheme
-		Logger            *zap.Logger
-		Informer          *informer.Manager
-		RunReconcileLocks sync.Map
-		WatcherStartLock  sync.Map
+		Scheme             *kRuntime.Scheme
+		Logger             *zap.Logger
+		Informer           *informer.Manager
+		SwitchModeLocks    sync.Map
+		InformerStartLocks sync.Map
+		ReconcileLocks     sync.Map
 	}
 )
 
 const (
-	ServeMode = "serve"
-	ProxyMode = "proxy"
-	NullMode  = ""
 
 	// These are resolver details, ideally in future we can move this to a configmap, or find a better way to serve this
 	resolverNamespace      = "elasti"
@@ -69,6 +67,13 @@ func (r *ElastiServiceReconciler) Reconcile(ctx context.Context, req ctrl.Reques
 			r.Logger.Error("Panic stack trace", zap.ByteString("stacktrace", buf[:n]))
 		}
 	}()
+
+	r.Logger.Debug("- In Reconcile", zap.String("es", req.NamespacedName.String()))
+	mutex := r.getMutexForReconcile(req.NamespacedName.String())
+	mutex.Lock()
+	defer r.Logger.Debug("- Out of Reconcile", zap.String("es", req.NamespacedName.String()))
+	defer mutex.Unlock()
+
 	// First we get the ElastiService object
 	// No mutex is taken for this, as we are not modifying the object, but if we face issues in future, we can add a mutex
 	es, esErr := r.getCRD(ctx, req.NamespacedName)
@@ -80,12 +85,6 @@ func (r *ElastiServiceReconciler) Reconcile(ctx context.Context, req ctrl.Reques
 		r.Logger.Error("Failed to get ElastiService in Reconcile", zap.String("es", req.String()), zap.Error(esErr))
 		return res, esErr
 	}
-	// We add the CRD details to service directory, so when elasti server received a request,
-	// we can find the right resource to scale up
-	crdDirectory.CRDDirectory.AddCRD(es.Spec.Service, &crdDirectory.CRDDetails{
-		CRDName:        es.Name,
-		DeploymentName: es.Spec.DeploymentName,
-	})
 
 	// If the ElastiService is being deleted, we need to clean up the resources
 	if !es.ObjectMeta.DeletionTimestamp.IsZero() {
@@ -103,25 +102,30 @@ func (r *ElastiServiceReconciler) Reconcile(ctx context.Context, req ctrl.Reques
 		return res, nil
 	}
 
-	// We check if the CRD is being deleted, and if it is, we clean up the resources
 	// We also check if the CRD has finalizer, and if not, we add the finalizer
 	if err := r.checkFinalizerCRD(ctx, es, req); err != nil {
 		r.Logger.Error("Failed to finalize CRD", zap.String("es", req.String()), zap.Error(err))
 		return res, err
 	}
 
+	// We add the CRD details to service directory, so when elasti server received a request,
+	// we can find the right resource to scale up
+	crdDirectory.CRDDirectory.AddCRD(es.Spec.Service, &crdDirectory.CRDDetails{
+		CRDName: es.Name,
+		Spec:    es.Spec,
+	})
+
 	// We need to start the informer only once per CRD. This is to avoid multiple informers for the same CRD
 	// We reset mutex if crd is deleted, so it can be used again if the same CRD is reapplied
 	r.getMutexForInformerStart(req.NamespacedName.String()).Do(func() {
-		// Watch for changes in target deployment
-		//go r.Informer.AddDeploymentWatch(req, es.Spec.DeploymentName, req.Namespace, r.getTargetDeploymentChangeHandler(ctx, es, req))
-		// Watch for changes in ScaleTargetRef
 		targetGroup, targetVersion, err := utils.ParseAPIVersion(es.Spec.ScaleTargetRef.APIVersion)
 		if err != nil {
 			r.Logger.Error("Failed to parse API version", zap.String("APIVersion", es.Spec.ScaleTargetRef.APIVersion), zap.Error(err))
 			return
 		}
-		go r.Informer.Add(&informer.RequestWatch{
+
+		// Watch for changes in ScaleTargetRef
+		r.Informer.Add(&informer.RequestWatch{
 			Req:               req,
 			ResourceName:      es.Spec.ScaleTargetRef.Name,
 			ResourceNamespace: req.Namespace,
@@ -132,8 +136,9 @@ func (r *ElastiServiceReconciler) Reconcile(ctx context.Context, req ctrl.Reques
 			},
 			Handlers: r.getScaleTargetRefChangeHandler(ctx, es, req),
 		})
+
 		// Watch for changes in public service
-		go r.Informer.Add(&informer.RequestWatch{
+		r.Informer.Add(&informer.RequestWatch{
 			Req:               req,
 			ResourceName:      es.Spec.Service,
 			ResourceNamespace: es.Namespace,
@@ -144,6 +149,11 @@ func (r *ElastiServiceReconciler) Reconcile(ctx context.Context, req ctrl.Reques
 			},
 			Handlers: r.getPublicServiceChangeHandler(ctx, es, req),
 		})
+
+		r.Logger.Info("ScaleTargetRef and Public Service added to informer", zap.String("es", req.String()),
+			zap.String("scaleTargetRef", es.Spec.ScaleTargetRef.Name),
+			zap.String("public service", es.Spec.Service),
+		)
 	})
 
 	return res, nil
@@ -154,3 +164,8 @@ func (r *ElastiServiceReconciler) SetupWithManager(mgr ctrl.Manager) error {
 		For(&v1alpha1.ElastiService{}).
 		Complete(r)
 }
+
+func (r *ElastiServiceReconciler) getMutexForReconcile(key string) *sync.Mutex {
+	l, _ := r.ReconcileLocks.LoadOrStore(key, &sync.Mutex{})
+	return l.(*sync.Mutex)
+}