Merge pull request #20 from stefanprodan/scheduler

Add canary analysis schedule interval to CRD

README.md

@@ -25,8 +25,7 @@ helm repo add flagger https://flagger.app
 # install or upgrade
 helm upgrade -i flagger flagger/flagger \
 --namespace=istio-system \
---set metricsServer=http://prometheus.istio-system:9090 \
---set controlLoopInterval=1m
+--set metricsServer=http://prometheus.istio-system:9090 
 ```
 
 Flagger is compatible with Kubernetes >1.10.0 and Istio >1.0.0.
@@ -75,7 +74,7 @@ You can change the canary analysis _max weight_ and the _step weight_ percentage
 For a deployment named _podinfo_, a canary promotion can be defined using Flagger's custom resource:
 
 ```yaml
-apiVersion: flagger.app/v1alpha2
+apiVersion: flagger.app/v1alpha3
 kind: Canary
 metadata:
   name: podinfo
@@ -102,8 +101,10 @@ spec:
     - public-gateway.istio-system.svc.cluster.local
     # Istio virtual service host names (optional)
     hosts:
-    - app.iowa.weavedx.com
+    - podinfo.example.com
   canaryAnalysis:
+    # schedule interval (default 60s)
+    interval: 1m
     # max number of failed metric checks before rollback
     threshold: 10
     # max traffic percentage routed to canary

artifacts/canaries/canary.yaml

@@ -1,4 +1,4 @@
-apiVersion: flagger.app/v1alpha2
+apiVersion: flagger.app/v1alpha3
 kind: Canary
 metadata:
   name: podinfo
@@ -27,6 +27,8 @@ spec:
     hosts:
     - app.iowa.weavedx.com
   canaryAnalysis:
+    # schedule interval (default 60s)
+    interval: 10s
     # max number of failed metric checks before rollback
     threshold: 10
     # max traffic percentage routed to canary
@@ -50,7 +52,7 @@ spec:
     # external checks (optional)
     webhooks:
       - name: integration-tests
-        url: http://podinfo.test:9898/echo
+        url: https://httpbin.org/post
         timeout: 1m
         metadata:
           test: "all"

artifacts/flagger/crd.yaml

@@ -4,11 +4,14 @@ metadata:
   name: canaries.flagger.app
 spec:
   group: flagger.app
-  version: v1alpha2
+  version: v1alpha3
   versions:
-    - name: v1alpha2
+    - name: v1alpha3
       served: true
       storage: true
+    - name: v1alpha2
+      served: true
+      storage: false
     - name: v1alpha1
       served: true
       storage: false
@@ -39,7 +42,9 @@ spec:
                 name:
                   type: string
             autoscalerRef:
-              type: object
+              anyOf:
+                - type: string
+                - type: object
               required: ['apiVersion', 'kind', 'name']
               properties:
                 apiVersion:
@@ -56,6 +61,9 @@ spec:
                   type: number
             canaryAnalysis:
               properties:
+                interval:
+                  type: string
+                  pattern: "^[0-9]+(m|s)"
                 threshold:
                   type: number
                 maxWeight:
@@ -73,7 +81,7 @@ spec:
                           type: string
                         interval:
                           type: string
-                          pattern: "^[0-9]+(m)"
+                          pattern: "^[0-9]+(m|s)"
                         threshold:
                           type: number
                 webhooks:
@@ -90,4 +98,4 @@ spec:
                           format: url
                         timeout:
                           type: string
-                          pattern: "^[0-9]+(s)"
+                          pattern: "^[0-9]+(m|s)"

artifacts/flagger/deployment.yaml

@@ -22,7 +22,7 @@ spec:
       serviceAccountName: flagger
       containers:
       - name: flagger
-        image: quay.io/stefanprodan/flagger:0.2.0
+        image: quay.io/stefanprodan/flagger:0.3.0-beta.1
         imagePullPolicy: Always
         ports:
         - name: http

charts/flagger/Chart.yaml

@@ -1,11 +1,11 @@
 apiVersion: v1
 name: flagger
 version: 0.2.0
-appVersion: 0.2.0
+appVersion: 0.3.0-beta.1
 kubeVersion: ">=1.9.0-0"
 engine: gotpl
 description: Flagger is a Kubernetes operator that automates the promotion of canary deployments using Istio routing for traffic shifting and Prometheus metrics for canary analysis.
-home: https://flagger.app
+home: https://docs.flagger.app
 icon: https://raw.githubusercontent.com/stefanprodan/flagger/master/docs/logo/flagger-icon.png
 sources:
 - https://github.com/stefanprodan/flagger

charts/flagger/README.md

@@ -1,7 +1,7 @@
 # Flagger
 
-[Flagger](https://flagger.app) is a Kubernetes operator that automates the promotion of canary deployments
-using Istio routing for traffic shifting and Prometheus metrics for canary analysis. 
+[Flagger](https://github.com/stefanprodan/flagger) is a Kubernetes operator that automates the promotion of 
+canary deployments using Istio routing for traffic shifting and Prometheus metrics for canary analysis. 
 Flagger implements a control loop that gradually shifts traffic to the canary while measuring key performance indicators
 like HTTP requests success rate, requests average duration and pods health.
 Based on the KPIs analysis a canary is promoted or aborted and the analysis result is published to Slack.
@@ -48,7 +48,6 @@ Parameter | Description | Default
 `image.repository` | image repository | `quay.io/stefanprodan/flagger`
 `image.tag` | image tag | `<VERSION>`
 `image.pullPolicy` | image pull policy | `IfNotPresent`
-`controlLoopInterval` | wait interval between checks | `10s`
 `metricsServer` | Prometheus URL | `http://prometheus.istio-system:9090`
 `slack.url` | Slack incoming webhook | None
 `slack.channel` | Slack channel | None
@@ -68,7 +67,8 @@ Specify each parameter using the `--set key=value[,key=value]` argument to `helm
 ```console
 $ helm upgrade -i flagger flagger/flagger \
   --namespace istio-system \
-  --set controlLoopInterval=1m
+  --set slack.url=https://hooks.slack.com/services/YOUR/SLACK/WEBHOOK \
+  --set slack.channel=general
 ```
 
 Alternatively, a YAML file that specifies the values for the above parameters can be provided while installing the chart. For example,
@@ -80,5 +80,5 @@ $ helm upgrade -i flagger flagger/flagger \
 ```
 
 > **Tip**: You can use the default [values.yaml](values.yaml)
-```
+
 

charts/flagger/templates/crd.yaml

@@ -5,11 +5,14 @@ metadata:
   name: canaries.flagger.app
 spec:
   group: flagger.app
-  version: v1alpha2
+  version: v1alpha3
   versions:
-    - name: v1alpha2
+    - name: v1alpha3
       served: true
       storage: true
+    - name: v1alpha2
+      served: true
+      storage: false
     - name: v1alpha1
       served: true
       storage: false
@@ -40,7 +43,9 @@ spec:
                 name:
                   type: string
             autoscalerRef:
-              type: object
+              anyOf:
+                - type: string
+                - type: object
               required: ['apiVersion', 'kind', 'name']
               properties:
                 apiVersion:
@@ -57,6 +62,9 @@ spec:
                   type: number
             canaryAnalysis:
               properties:
+                interval:
+                  type: string
+                  pattern: "^[0-9]+(m|s)"
                 threshold:
                   type: number
                 maxWeight:
@@ -74,7 +82,7 @@ spec:
                           type: string
                         interval:
                           type: string
-                          pattern: "^[0-9]+(m)"
+                          pattern: "^[0-9]+(m|s)"
                         threshold:
                           type: number
                 webhooks:
@@ -91,6 +99,5 @@ spec:
                           format: url
                         timeout:
                           type: string
-                          pattern: "^[0-9]+(s)"
-
+                          pattern: "^[0-9]+(m|s)"
 {{- end }}

charts/flagger/templates/deployment.yaml

@@ -35,7 +35,6 @@ spec:
           command:
           - ./flagger
           - -log-level=info
-          - -control-loop-interval={{ .Values.controlLoopInterval }}
           - -metrics-server={{ .Values.metricsServer }}
           {{- if .Values.slack.url }}
           - -slack-url={{ .Values.slack.url }}

charts/flagger/values.yaml

@@ -2,10 +2,9 @@
 
 image:
   repository: quay.io/stefanprodan/flagger
-  tag: 0.2.0
+  tag: 0.3.0-beta.1
   pullPolicy: IfNotPresent
 
-controlLoopInterval: "1m"
 metricsServer: "http://prometheus.istio-system.svc.cluster.local:9090"
 
 slack:

cmd/flagger/main.go

@@ -37,7 +37,7 @@ func init() {
 	flag.StringVar(&kubeconfig, "kubeconfig", "", "Path to a kubeconfig. Only required if out-of-cluster.")
 	flag.StringVar(&masterURL, "master", "", "The address of the Kubernetes API server. Overrides any value in kubeconfig. Only required if out-of-cluster.")
 	flag.StringVar(&metricsServer, "metrics-server", "http://prometheus:9090", "Prometheus URL")
-	flag.DurationVar(&controlLoopInterval, "control-loop-interval", 10*time.Second, "wait interval between rollouts")
+	flag.DurationVar(&controlLoopInterval, "control-loop-interval", 10*time.Second, "Kubernetes API sync interval")
 	flag.StringVar(&logLevel, "log-level", "debug", "Log level can be: debug, info, warning, error.")
 	flag.StringVar(&port, "port", "8080", "Port to listen on.")
 	flag.StringVar(&slackURL, "slack-url", "", "Slack hook URL.")
@@ -77,7 +77,7 @@ func main() {
 	}
 
 	flaggerInformerFactory := informers.NewSharedInformerFactory(flaggerClient, time.Second*30)
-	canaryInformer := flaggerInformerFactory.Flagger().V1alpha2().Canaries()
+	canaryInformer := flaggerInformerFactory.Flagger().V1alpha3().Canaries()
 
 	logger.Infof("Starting flagger version %s revision %s", version.VERSION, version.REVISION)
 

docs/gitbook/how-it-works.md

@@ -9,7 +9,7 @@
 For a deployment named _podinfo_, a canary promotion can be defined using Flagger's custom resource:
 
 ```yaml
-apiVersion: flagger.app/v1alpha2
+apiVersion: flagger.app/v1alpha3
 kind: Canary
 metadata:
   name: podinfo
@@ -38,6 +38,8 @@ spec:
     hosts:
     - podinfo.example.com
   canaryAnalysis:
+    # schedule interval (default 60s)
+    interval: 1m
     # max number of failed metric checks before rollback
     threshold: 10
     # max traffic percentage routed to canary
@@ -128,12 +130,13 @@ You can change the canary analysis _max weight_ and the _step weight_ percentage
 ### Canary Analysis
 
 The canary analysis runs periodically until it reaches the maximum traffic weight or the failed checks threshold. 
-By default the analysis interval is set to one minute and can be configured with the `controlLoopInterval` command flag.
 
 Spec:
 
 ```yaml
   canaryAnalysis:
+    # schedule interval (default 60s)
+    interval: 1m
     # max number of failed metric checks before rollback
     threshold: 10
     # max traffic percentage routed to canary
@@ -148,13 +151,13 @@ The above analysis, if it succeeds, will run for 25 minutes while validating the
 You can determine the minimum time that it takes to validate and promote a canary deployment using this formula:
 
 ```
-controlLoopInterval * (maxWeight / stepWeight)
+interval * (maxWeight / stepWeight)
 ```
 
-And the time it takes for a canary to be rollback:
+And the time it takes for a canary to be rollback when the metrics or webhook checks are failing:
 
 ```
-controlLoopInterval * threshold 
+interval * threshold 
 ```
 
 ### HTTP Metrics

docs/gitbook/install/install-flagger.md

@@ -23,8 +23,7 @@ Deploy Flagger in the _**istio-system**_ namespace:
 ```bash
 helm upgrade -i flagger flagger/flagger \
 --namespace=istio-system \
---set metricsServer=http://prometheus.istio-system:9090 \
---set controlLoopInterval=1m
+--set metricsServer=http://prometheus.istio-system:9090
 ```
 
 Enable **Slack** notifications:

docs/gitbook/usage/progressive-delivery.md

@@ -20,7 +20,7 @@ kubectl apply -f ${REPO}/artifacts/canaries/hpa.yaml
 Create a canary custom resource \(replace example.com with your own domain\):
 
 ```yaml
-apiVersion: v1alpha2
+apiVersion: flagger.app/v1alpha3
 kind: Canary
 metadata:
   name: podinfo
@@ -49,6 +49,8 @@ spec:
     hosts:
     - app.example.com
   canaryAnalysis:
+    # schedule interval (default 60s)
+    interval: 1m
     # max number of failed metric checks before rollback
     threshold: 5
     # max traffic percentage routed to canary

hack/update-codegen.sh

@@ -23,6 +23,6 @@ CODEGEN_PKG=${CODEGEN_PKG:-$(cd ${SCRIPT_ROOT}; ls -d -1 ./vendor/k8s.io/code-ge
 
 ${CODEGEN_PKG}/generate-groups.sh "deepcopy,client,informer,lister" \
   github.com/stefanprodan/flagger/pkg/client github.com/stefanprodan/flagger/pkg/apis \
-  flagger:v1alpha2 \
+  flagger:v1alpha3 \
   --go-header-file ${SCRIPT_ROOT}/hack/boilerplate.go.txt
 

pkg/apis/flagger/v1alpha2/doc.go → pkg/apis/flagger/v1alpha3/doc.go

@@ -16,6 +16,6 @@ limitations under the License.
 
 // +k8s:deepcopy-gen=package
 
-// Package v1alpha2 is the v1alpha2 version of the API.
+// Package v1alpha3 is the v1alpha3 version of the API.
 // +groupName=flagger.app
-package v1alpha2
+package v1alpha3

pkg/apis/flagger/v1alpha2/register.go → pkg/apis/flagger/v1alpha3/register.go

@@ -14,7 +14,7 @@ See the License for the specific language governing permissions and
 limitations under the License.
 */
 
-package v1alpha2
+package v1alpha3
 
 import (
 	metav1 "k8s.io/apimachinery/pkg/apis/meta/v1"
@@ -25,7 +25,7 @@ import (
 )
 
 // SchemeGroupVersion is group version used to register these objects
-var SchemeGroupVersion = schema.GroupVersion{Group: rollout.GroupName, Version: "v1alpha2"}
+var SchemeGroupVersion = schema.GroupVersion{Group: rollout.GroupName, Version: "v1alpha3"}
 
 // Kind takes an unqualified kind and returns back a Group qualified GroupKind
 func Kind(kind string) schema.GroupKind {