Add support for StatefulSets

README.md

@@ -12,14 +12,14 @@ development.__
 
 ## How it works
 
-For Osiris-enabled deployments, Osiris automatically instruments application
-pods with a __metrics-collecting proxy__ deployed as a sidecar container.
+For Osiris-enabled deployments or statefulSets, Osiris automatically instruments
+application pods with a __metrics-collecting proxy__ deployed as a sidecar container.
 
-For any Osiris-enabled deployment that is _already_ scaled to a configurable
-minimum number of replicas (one, by default), the __zeroscaler__ component
-continuously analyzes metrics from each of that deployment's pods. When the
-aggregated metrics reveal that all of the deployment's pods are idling, the
-zeroscaler scales the deployment to zero replicas.
+For any Osiris-enabled deployment or statefulSet that is _already_ scaled to
+a configurable minimum number of replicas (one, by default), the __zeroscaler__
+component continuously analyzes metrics from each of that deployment/statefulSet's
+pods. When the aggregated metrics reveal that all of the deployment/statefulSet's
+pods are idling, the zeroscaler scales the deployment/statefulSet to zero replicas.
 
 Under normal circumstances, scaling a deployment to zero replicas poses a
 problem: any services that select pods from that deployment (and only that
@@ -170,14 +170,14 @@ spec:
 
 Most of Osiris configuration is done with Kubernetes annotations - as seen in the Usage section.
 
-#### Deployment Annotations
+#### Deployment & StatefulSet Annotations
 
-The following table lists the supported annotations for Kubernetes `Deployments` and their default values.
+The following table lists the supported annotations for Kubernetes `Deployments` and `StatefulSets`, and their default values.
 
 | Annotation | Description | Default |
 | ---------- | ----------- | ------- |
-| `osiris.deislabs.io/enabled` | Enable Osiris for this deployment. Allowed values: `y`, `yes`, `true`, `on`, `1`. | _no value_ (= disabled) |
-| `osiris.deislabs.io/minReplicas` | The minimum number of replicas to set on the deployment when Osiris will scale up. If you set `2`, Osiris will scale the deployment from `0` to `2` replicas directly. Osiris won't collect metrics from deployments which have more than `minReplicas` replicas - to avoid useless collections of metrics. | `1` |
+| `osiris.deislabs.io/enabled` | Enable Osiris for this deployment or statefulSet. Allowed values: `y`, `yes`, `true`, `on`, `1`. | _no value_ (= disabled) |
+| `osiris.deislabs.io/minReplicas` | The minimum number of replicas to set on the deployment/statefulSet when Osiris will scale up. If you set `2`, Osiris will scale the deployment/statefulSet from `0` to `2` replicas directly. Osiris won't collect metrics from deployments/statefulSets which have more than `minReplicas` replicas - to avoid useless collections of metrics. | `1` |
 
 #### Service Annotations
 
@@ -187,6 +187,7 @@ The following table lists the supported annotations for Kubernetes `Services` an
 | ---------- | ----------- | ------- |
 | `osiris.deislabs.io/enabled` | Enable Osiris for this service. Allowed values: `y`, `yes`, `true`, `on`, `1`. | _no value_ (= disabled) |
 | `osiris.deislabs.io/deployment` | Name of the deployment which is behind this service. This is required to map the service with its deployment. | _no value_ |
+| `osiris.deislabs.io/statefulset` | Name of the statefulSet which is behind this service. This is required to map the service with its statefulSet. | _no value_ |
 | `osiris.deislabs.io/loadBalancerHostname` | Map requests coming from a specific hostname to this service. Note that if you have multiple hostnames, you can set them with different annotations, using `osiris.deislabs.io/loadBalancerHostname-1`, `osiris.deislabs.io/loadBalancerHostname-2`, ... | _no value_ |
 | `osiris.deislabs.io/ingressHostname` | Map requests coming from a specific hostname to this service. If you use an ingress in front of your service, this is required to create a link between the ingress and the service. Note that if you have multiple hostnames, you can set them with different annotations, using `osiris.deislabs.io/ingressHostname-1`, `osiris.deislabs.io/ingressHostname-2`, ... | _no value_ |
 | `osiris.deislabs.io/ingressDefaultPort` | Custom service port when the request comes from an ingress. Default behaviour if there are more than 1 port on the service, is to look for a port named `http`, and fallback to the port `80`. Set this if you have multiple ports and using a non-standard port with a non-standard name. | _no value_ |
@@ -199,7 +200,7 @@ Note that you might see an `osiris.deislabs.io/selector` annotation - this is fo
 Deploy the [example application](example/hello-osiris.yaml) `hello-osiris` :
 
 ```
-kubectl create -f ./example/hello-osiris.yaml
+kubectl create -f ./example/hello-osiris-deployment.yaml
 ```
 
 This will create an Osiris-enabled deployment and service named `hello-osiris`.
@@ -219,6 +220,10 @@ replicas and the one `hello-osiris` pod should be terminated.
 Make a request again, and watch as Osiris scales the deployment back to one
 replica and your request is handled successfully.
 
+You can also do the same with the 
+[hello-osiris-statefulset.yaml](example/hello-osiris-statefulset.yaml) example, 
+which uses a Kubernetes StatefulSet instead of a Deployment.
+
 ## Limitations
 
 It is a specific goal of Osiris to enable greater resource efficiency within

chart/osiris/templates/cluster-role.yaml

@@ -32,6 +32,7 @@ rules:
   - apps
   resources:
   - deployments
+  - statefulsets
   verbs:
   - get
   - list

chart/osiris/templates/proxy-injector-webhook-config.yaml

@@ -42,6 +42,7 @@ webhooks:
     - v1
     resources:
     - deployments
+    - statefulsets
     operations:
     - CREATE
     - UPDATE

example/hello-osiris-statefulset.yaml

@@ -0,0 +1,87 @@
+apiVersion: v1
+kind: Service
+metadata:
+  name: hello-osiris
+  labels:
+    app: hello-osiris
+  annotations:
+    osiris.deislabs.io/enabled: "true"
+    osiris.deislabs.io/statefulset: hello-osiris
+    osiris.deislabs.io/loadBalancerHostname: hello-osiris.contoso.io
+spec:
+  type: LoadBalancer
+  ports:
+  - name: http
+    port: 8080
+    targetPort: 8080
+  - name: h2c
+    port: 8081
+    targetPort: 8081
+  - name: grpc
+    port: 8082
+    targetPort: 8082
+  - name: https
+    port: 4430
+    targetPort: 4430
+  selector:
+    app: hello-osiris
+---
+apiVersion: v1
+kind: Secret
+metadata:
+  name: hello-osiris-cert
+  labels:
+    app: hello-osiris
+data:
+  server.crt: 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
+  server.key: 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
+---
+apiVersion: apps/v1
+kind: StatefulSet
+metadata:
+  name: hello-osiris
+  labels:
+    app: hello-osiris
+  annotations:
+    osiris.deislabs.io/enabled: "true"
+    osiris.deislabs.io/minReplicas: "1"
+spec:
+  replicas: 1
+  serviceName: hello-osiris
+  selector:
+    matchLabels:
+      app: hello-osiris
+  template:
+    metadata:
+      labels:
+        app: hello-osiris
+    spec:
+      containers:
+      - name: hello-osiris
+        image: krancour/hello-osiris:v0.1.0
+        args:
+        - --https-cert
+        - /hello-osiris/cert/server.crt
+        - --https-key
+        - /hello-osiris/cert/server.key
+        ports:
+        - containerPort: 8080
+        - containerPort: 8081
+        - containerPort: 8082
+        - containerPort: 4430
+        volumeMounts:
+        - name: cert
+          mountPath: /hello-osiris/cert
+          readOnly: true
+        livenessProbe:
+          httpGet:
+            path: /healthz
+            port: 8080
+        readinessProbe:
+          httpGet:
+            path: /healthz
+            port: 8080
+      volumes:
+      - name: cert
+        secret:
+          secretName: hello-osiris-cert

pkg/deployments/activator/activating.go

@@ -12,23 +12,23 @@ import (
 
 func (a *activator) activateDeployment(
 	app *app,
-) (*deploymentActivation, error) {
+) (*appActivation, error) {
 	deploymentsClient := a.kubeClient.AppsV1().Deployments(app.namespace)
 	deployment, err := deploymentsClient.Get(
-		app.deploymentName,
+		app.name,
 		metav1.GetOptions{},
 	)
 	if err != nil {
 		return nil, err
 	}
-	da := &deploymentActivation{
+	da := &appActivation{
 		readyAppPodIPs: map[string]struct{}{},
 		successCh:      make(chan struct{}),
 		timeoutCh:      make(chan struct{}),
 	}
 	glog.Infof(
 		"Activating deployment %s in namespace %s",
-		app.deploymentName,
+		app.name,
 		app.namespace,
 	)
 	go da.watchForCompletion(
@@ -50,7 +50,54 @@ func (a *activator) activateDeployment(
 	}}
 	patchesBytes, _ := json.Marshal(patches)
 	_, err = deploymentsClient.Patch(
-		app.deploymentName,
+		app.name,
+		k8s_types.JSONPatchType,
+		patchesBytes,
+	)
+	return da, err
+}
+
+func (a *activator) activateStatefulSet(
+	app *app,
+) (*appActivation, error) {
+	statefulSetsClient := a.kubeClient.AppsV1().StatefulSets(app.namespace)
+	statefulSet, err := statefulSetsClient.Get(
+		app.name,
+		metav1.GetOptions{},
+	)
+	if err != nil {
+		return nil, err
+	}
+	da := &appActivation{
+		readyAppPodIPs: map[string]struct{}{},
+		successCh:      make(chan struct{}),
+		timeoutCh:      make(chan struct{}),
+	}
+	glog.Infof(
+		"Activating statefulSet %s in namespace %s",
+		app.name,
+		app.namespace,
+	)
+	go da.watchForCompletion(
+		a.kubeClient,
+		app,
+		labels.Set(statefulSet.Spec.Selector.MatchLabels).AsSelector(),
+	)
+	if statefulSet.Spec.Replicas == nil || *statefulSet.Spec.Replicas > 0 {
+		// We don't need to do this, as it turns out! Scaling is either already
+		// in progress-- perhaps initiated by another process-- or may even be
+		// completed already. Just return dr and allow the caller to move on to
+		// verifying / waiting for this activation to be complete.
+		return da, nil
+	}
+	patches := []kubernetes.PatchOperation{{
+		Op:    "replace",
+		Path:  "/spec/replicas",
+		Value: kubernetes.GetMinReplicas(statefulSet.Annotations, 1),
+	}}
+	patchesBytes, _ := json.Marshal(patches)
+	_, err = statefulSetsClient.Patch(
+		app.name,
 		k8s_types.JSONPatchType,
 		patchesBytes,
 	)

pkg/deployments/activator/activator.go

@@ -29,8 +29,8 @@ type activator struct {
 	nodeAddresses             map[string]struct{}
 	appsByHost                map[string]*app
 	indicesLock               sync.RWMutex
-	deploymentActivations     map[string]*deploymentActivation
-	deploymentActivationsLock sync.Mutex
+	appActivations            map[string]*appActivation
+	appActivationsLock        sync.Mutex
 	dynamicProxyListenAddrStr string
 	dynamicProxy              tcp.DynamicProxy
 	httpClient                *http.Client
@@ -56,7 +56,7 @@ func NewActivator(kubeClient kubernetes.Interface) (Activator, error) {
 		services:                  map[string]*corev1.Service{},
 		nodeAddresses:             map[string]struct{}{},
 		appsByHost:                map[string]*app{},
-		deploymentActivations:     map[string]*deploymentActivation{},
+		appActivations:            map[string]*appActivation{},
 		httpClient: &http.Client{
 			Timeout: time.Minute * 1,
 		},
@@ -127,7 +127,7 @@ func (a *activator) syncService(obj interface{}) {
 	a.indicesLock.Lock()
 	defer a.indicesLock.Unlock()
 	svc := obj.(*corev1.Service)
-	svcKey := getKey(svc.Namespace, svc.Name)
+	svcKey := getKey(svc.Namespace, "Service", svc.Name)
 	if k8s.ResourceIsOsirisEnabled(svc.Annotations) {
 		a.services[svcKey] = svc
 	} else {
@@ -140,7 +140,7 @@ func (a *activator) syncDeletedService(obj interface{}) {
 	a.indicesLock.Lock()
 	defer a.indicesLock.Unlock()
 	svc := obj.(*corev1.Service)
-	svcKey := getKey(svc.Namespace, svc.Name)
+	svcKey := getKey(svc.Namespace, "Service", svc.Name)
 	delete(a.services, svcKey)
 	a.updateIndex()
 }

pkg/deployments/activator/app.go

@@ -1,9 +1,17 @@
 package activator
 
+type appKind string
+
+const (
+	appKindDeployment  appKind = "Deployment"
+	appKindStatefulSet appKind = "StatefulSet"
+)
+
 type app struct {
-	namespace      string
-	serviceName    string
-	deploymentName string
-	targetHost     string
-	targetPort     int
+	namespace   string
+	serviceName string
+	name        string
+	kind        appKind
+	targetHost  string
+	targetPort  int
 }