v1.0.0 Release: General Availability

Makefile

@@ -7,8 +7,14 @@ UNIQUE:=$(shell date +%s)
 # Where to push the docker image.
 REGISTRY := practodev
 
+BASE_BRANCH := master
+CURRENT_BRANCH := $(shell git rev-parse --abbrev-ref HEAD)
 # This version-strategy uses git tags to set the version string
 VERSION := $(shell git describe --tags --always --dirty)
+MAJOR_VERSION = $(shell git describe --tags  --dirty | \
+	awk -F'.' '{print $$1}')
+MAJOR_MINOR_VERSION = $(shell git describe --tags  --dirty | \
+	awk -F'.' '{print $$1"."$$2}')
 #
 # This version-strategy uses a manual value to set the version string
 #VERSION := 1.2.3
@@ -30,13 +36,23 @@ ARCH := $(if $(GOARCH),$(GOARCH),$(shell go env GOARCH))
 BASEIMAGE ?= gcr.io/distroless/static
 
 IMAGE := $(REGISTRY)/$(BIN)
+
 TAG := $(VERSION)
-BETA := v1.0-beta
+MAJOR_TAG := $(MAJOR_VERSION)
+MAJOR_MINOR_TAG := $(MAJOR_MINOR_VERSION)
 
-ifneq (,$(findstring v1.0,$(TAG))$(findstring -beta,$(TAG)))
-    PUBLISH_TAGS := $(IMAGE):$(TAG) $(IMAGE):$(BETA)
-else
-    PUBLISH_TAGS := $(IMAGE):$(TAG)
+PUBLISH_TAGS := $(IMAGE):$(TAG) $(IMAGE):$(MAJOR_MINOR_VERSION) $(IMAGE):$(MAJOR_VERSION)
+ifneq (,$(findstring -beta,$(TAG)))
+    PUBLISH_TAGS := $(IMAGE):$(TAG) $(IMAGE):$(MAJOR_MINOR_VERSION)-beta $(IMAGE):$(MAJOR_VERSION)-beta
+endif
+ifneq (,$(findstring -alpha,$(TAG)))
+    PUBLISH_TAGS := $(IMAGE):$(TAG) $(IMAGE):$(MAJOR_MINOR_VERSION)-alpha $(IMAGE):$(MAJOR_VERSION)-alpha
+endif
+ifneq ($(CURRENT_BRANCH), $(BASE_BRANCH))
+	PUBLISH_TAGS := $(IMAGE):$(TAG)
+endif
+ifneq (,$(findstring -dirty,$(TAG)))
+	PUBLISH_TAGS := $(IMAGE):$(TAG)
 endif
 
 define \n

README.md

@@ -12,14 +12,14 @@ Currently the supported Message Queueing Services are:
 - [AWS SQS](https://aws.amazon.com/sqs/)
 - [Beanstalkd](https://beanstalkd.github.io/)
 
-There is a plan to integrate other commonly used message queing services.
+There is a plan to integrate other commonly used message queuing services.
 
 ----
 
 # Install the WorkerPodAutoscaler
 
 ### Install
-Running the below script will create the WPA [CRD](https://kubernetes.io/docs/concepts/extend-kubernetes/api-extension/custom-resources/) and start the controller. The controller watches over all the specified queues in AWS SQS and scales the Kubernetes deployments based on the specification.
+Running the below script will create the WPA [CRD](https://kubernetes.io/docs/concepts/extend-kubernetes/api-extension/custom-resources/) and install the worker pod autoscaler deployment.
 
 ```bash
 export AWS_REGIONS='ap-south-1,ap-southeast-1'
@@ -28,7 +28,7 @@ export AWS_SECRET_ACCESS_KEY='sample-aws-secret-acesss-key'
 ./hack/install.sh
 ```
 
-Note: IAM policy required is [this](artifacts/iam-policy.json).
+Note: IAM policy required is [this](artifacts/iam-policy.json)
 
 ### Verify Installation
 Check the wpa resource is accessible using kubectl
@@ -37,8 +37,12 @@ Check the wpa resource is accessible using kubectl
 kubectl get wpa
 ```
 
+### Upgrade
+
+Please follow [this document](UPGRADE.md) for upgrading Worker Pod Autoscaler.
+
 # Example
-Do install the controller before going with the example.
+Do install the wpa crd and wpa deployment before going with the example. (Please check above.)
 
 - Create Deployment that needs to scale based on queue length.
 ```bash
@@ -56,7 +60,7 @@ This will start scaling `example-deployment` based on SQS queue length.
 
 ## WPA Resource
 ```yaml
-apiVersion: k8s.practo.dev/v1alpha1
+apiVersion: k8s.practo.dev/v1
 kind: WorkerPodAutoScaler
 metadata:
   name: example-wpa
@@ -71,14 +75,19 @@ spec:
 ```
 Beanstalk's queueURI would be like: `beanstalk://beanstalkDNSName:11300/test-tube`
 
-### WPA Spec documentation:
-- **minReplicas**: minimum number of workers you want to run. (mandatory)
-- **maxReplicas**: maximum number of workers you want to run. (mandatory)
-- **deploymentName**: name of the kubernetes deployment in the same namespace as WPA object. (mandatory)
-- **queueURI**: full URL of the queue. (mandatory)
-- **targetMessagesPerWorker**: Number of jobs in the queue which have not been picked up by the workers. This also used to calculate the desired number of workers. (mandatory)
-- **secondsToProcessOneJob:**: This metric is useful to calculate the desired number of workers more accurately. It is particularly very useful for workers which have `targetMessagesPerWorker` as always zero. `secondsToProcessOneJob` in the combination with `messagesSentPerMinute`(queue RPM) helps in calculating the minimum workers that is expected to be running to handle `messagesSentPerMinute`(RPM) with every job being processed in `secondsToProcessOneJob` seconds. (optional, highly recommended, default=0.0 i.e. disabled)
-- **maxDisruption:** amount of disruption that can be tolerated in a single scale down activity. Number of pods or percentage of pods that can scale down in a single down scale down activity. Using this you can control how fast a scale down can happen. This can be expressed both as an absolute value and a percentage. Explained with the help of some examples: (optional, default is the WPA flag `wpa-default-max-disruption`)
+### WPA Spec Documentation:
+
+| Spec          | Description   | Mandatory |
+| :------------ | :----------- |:------------|
+| minReplicas    | Minimum number of workers you want to run.                                | Yes |
+| maxReplicas    | Maximum number of workers you want to run                                 | Yes |
+| deploymentName | Name of the kubernetes deployment in the same namespace as WPA object. | Yes |
+| queueURI       | Full URL of the queue.                                                 | Yes |
+| targetMessagesPerWorker |  Number of jobs in the queue which have not been picked up by the workers. This is also used to calculate the desired number of workers. | Yes |
+| secondsToProcessOneJob | This metric is useful to calculate the desired number of workers more accurately. It is particularly very useful for workers which have `targetMessagesPerWorker` as always zero. `secondsToProcessOneJob` in the combination with `messagesSentPerMinute`(queue RPM) helps in calculating the minimum workers that is expected to be running to handle `messagesSentPerMinute`(RPM) with every job being processed in `secondsToProcessOneJob` seconds. (highly recommended, default=0.0 i.e. disabled). | No |
+| maxDisruption | Amount of disruption that can be tolerated in a single scale down activity. Number of pods or percentage of pods that can scale down in a single down scale down activity. Using this you can control how fast a scale down can happen. This can be expressed both as an absolute value and a percentage. (default is the WPA flag `--wpa-default-max-disruption`). | No |
+
+`maxDisruption` explained with the help of some examples:
 ```
 min=2, max=1000, current=500, maxDisruption=50%: then the scale down cannot bring down more than 250 pods in a single scale down activity.
 ```
@@ -127,7 +136,7 @@ If you need to enable multiple queue support, you can add queues comma separated
 
 Running WPA at scale require changes in `--k8s-api-burst` and `--k8s-api-qps` flags.
 
-WPA makes update to Kubernetes API to update the WPA status. WPA uses [client-go](https://github.com/kubernetes/client-go) as the kubernetes client to make Kubernetes API calls. This client allows 5QPS and 10Burst requests to Kubernetes API by default. The defaults can be changed by using `k8s-api-burst` and `k8s-api-qps` flags.
+WPA makes call to the Kubernetes API to update the WPA resource status. [client-go](https://github.com/kubernetes/client-go) is used as the kubernetes client to make the Kubernetes API calls. This client allows 5QPS and 10Burst requests to Kubernetes API by default. The defaults can be changed by using `k8s-api-burst` and `k8s-api-qps` flags.
 
 You may need to increase the `--k8s-api-qps` and `k8s-api-burst` if [wpa_controller_loop_duration_seconds](https://github.com/practo/k8s-worker-pod-autoscaler/tree/master#wpa-metrics) is greater than 200ms (wpa_controller_loop_duration_seconds>0.200)
 
@@ -183,11 +192,12 @@ git pull origin master
 - Build and push the image to `hub.docker.com/practodev`. Note: practodev push access is required.
 ```
 git fetch --tags
-git tag v1.0.0-beta
+git tag v1.0.0
 make push
 ```
+Note: For every tag major and major minor versions tags also available. For example: `v1` and `v1.0`
 
-- Create a Release in Github. Refer this https://github.com/practo/k8s-worker-pod-autoscaler/releases/tag/v1.0.0-beta and create a release. Release should contain the Changelog information of all the issues and pull request after the last release.
+- Create a Release in Github. Refer [this](https://github.com/practo/k8s-worker-pod-autoscaler/releases/tag/v1.0.0) and create a release. Release should contain the Changelog information of all the issues and pull request after the last release.
 
 -  Publish the release in Github 🎉
 
@@ -199,7 +209,7 @@ kubectl edit deployment -n kube-system workerpodautoscaler
 ```
 
 ## Contributing
-It would be really helpful to add all the major message queuing service providers. This [interface](https://github.com/practo/k8s-worker-pod-autoscaler/blob/master/pkg/queue/queueing_service.go#L5-L14) implementation needs to be written down to make that possible.
+It would be really helpful to add all the major message queuing service providers. This [interface](https://github.com/practo/k8s-worker-pod-autoscaler/blob/master/pkg/queue/queue_service.go) implementation needs to be written down to make that possible.
 
 - After making code changes, run the below commands to build and run locally.
 ```
@@ -217,7 +227,7 @@ $ bin/darwin_amd64/workerpodautoscaler run --kube-config /home/user/.kube/config
 
 ## Thanks
 
-Thanks to kubernetes team for making [crds](https://kubernetes.io/docs/concepts/extend-kubernetes/api-extension/custom-resources/) and [sample controller](https://github.com/kubernetes/sample-controller)
+Thanks to kubernetes team for making [crds](https://kubernetes.io/docs/concepts/extend-kubernetes/api-extension/custom-resources/) and [sample controller](https://github.com/kubernetes/sample-controller). Thanks for [go-build-template](https://github.com/thockin/go-build-template).
 
 [latest-release]: https://github.com/practo/k8s-worker-pod-autoscaler/releases
 [GoDoc]: https://godoc.org/github.com/practo/k8s-worker-pod-autoscaler

UPGRADE.md

@@ -0,0 +1,18 @@
+# Upgrade Worker Pod Autoscaler
+
+## Upgrade from v0.2 to v1
+
+### Breaking Changes
+There is no backward breaking change from `v0.2` to `v1`.
+
+### Recommended Actions
+Update the WorkerPodAutoScaler CRD from `v1alpha1` to `v1` using below:
+```
+kubectl apply -f ./artifacts/crd.yaml
+```
+
+Note: Support for `v1alpha1` CRD version is still there, but it would be discontinued in the future releases.
+
+### Changes
+- [v1.0.0](https://github.com/practo/k8s-worker-pod-autoscaler/releases/tag/v1.0.0)
+- [v1.0.0-beta](https://github.com/practo/k8s-worker-pod-autoscaler/releases/tag/v1.0.0-beta)

artifacts/crd.yaml

@@ -0,0 +1,74 @@
+apiVersion: apiextensions.k8s.io/v1beta1
+kind: CustomResourceDefinition
+metadata:
+  name: workerpodautoscalers.k8s.practo.dev
+spec:
+  conversion:
+    strategy: None
+  group: k8s.practo.dev
+  names:
+    kind: WorkerPodAutoScaler
+    listKind: WorkerPodAutoScalerList
+    plural: workerpodautoscalers
+    shortNames:
+    - wpa
+    - wpas
+    singular: workerpodautoscaler
+  preserveUnknownFields: true
+  scope: Namespaced
+  versions: v1
+  versions:
+  - name: v1
+    served: true
+    storage: true
+    schema:
+      openAPIV3Schema:
+        type: object
+        required:
+        - spec
+        properties:
+          apiVersion:
+            type: string
+          kind:
+            type: string
+          metadata:
+            type: object
+          spec:
+            type: object
+            required:
+            - deploymentName
+            - minReplicas
+            - maxReplicas
+            - queueURI
+            - targetMessagesPerWorker
+            properties:
+              deploymentName:
+                type: string
+                description: 'Name of the kubernetes deployment in the same namespace as WPA object'
+              maxDisruption:
+                type: string
+                nullable: true
+                description: 'Amount of disruption that can be tolerated in a single scale down activity. Number of pods or percentage of pods that can scale down in a single down scale down activity'
+              maxReplicas:
+                type: integer
+                format: int32
+                description: 'Maximum number of workers you want to run'
+              minReplicas:
+                type: integer
+                format: int32
+                description: 'Minimum number of workers you want to run'
+              queueURI:
+                type: string
+                description: 'Full URL of the queue'
+              targetMessagesPerWorker:
+                type: integer
+                format: int32
+                description: 'Number of jobs in the queue which have not been picked up by the workers. This also used to calculate the desired number of workers'
+              secondsToProcessOneJob:
+                type: number
+                format: float
+                nullable: true
+                description: 'This metric is useful to calculate the desired number of workers more accurately. It is particularly very useful for workers which have `targetMessagesPerWorker` as always zero. `secondsToProcessOneJob` in the combination with `messagesSentPerMinute`(queue RPM) helps in calculating the minimum workers that is expected to be running to handle `messagesSentPerMinute`(RPM) with every job being processed in `secondsToProcessOneJob` seconds'
+  - name: v1alpha1
+    served: true
+    storage: false

artifacts/deployment-template.yaml

@@ -28,7 +28,7 @@ spec:
          value: {{ WPA_AWS_ACCESS_KEY_ID }}
        - name: AWS_SECRET_ACCESS_KEY
          value: {{ WPA_AWS_SECRET_ACCESS_KEY }}
-       image: practodev/workerpodautoscaler:v1.0.0-beta
+       image: practodev/workerpodautoscaler:{{ WPA_TAG }}
        imagePullPolicy: Always
        command:
        - /workerpodautoscaler

artifacts/examples/example-wpa.yaml

@@ -1,4 +1,4 @@
-apiVersion: k8s.practo.dev/v1alpha1
+apiVersion: k8s.practo.dev/v1
 kind: WorkerPodAutoScaler
 metadata:
   name: example-wpa

cmd/workerpodautoscaler/run.go

@@ -14,7 +14,6 @@ import (
 	"k8s.io/client-go/rest"
 	"k8s.io/client-go/tools/clientcmd"
 
-	"github.com/practo/k8s-worker-pod-autoscaler/pkg/apis/workerpodautoscaler/v1alpha1"
 	"github.com/practo/k8s-worker-pod-autoscaler/pkg/cmdutil"
 	"github.com/practo/k8s-worker-pod-autoscaler/pkg/signals"
 	"github.com/prometheus/client_golang/prometheus/promhttp"
@@ -24,7 +23,6 @@ import (
 	clientset "github.com/practo/k8s-worker-pod-autoscaler/pkg/generated/clientset/versioned"
 	informers "github.com/practo/k8s-worker-pod-autoscaler/pkg/generated/informers/externalversions"
 	queue "github.com/practo/k8s-worker-pod-autoscaler/pkg/queue"
-	apiextensionsclient "k8s.io/apiextensions-apiserver/pkg/client/clientset/clientset"
 )
 
 type runCmd struct {
@@ -137,16 +135,6 @@ func (v *runCmd) run(cmd *cobra.Command, args []string) {
 		klog.Fatalf("Error building custom clientset: %s", err.Error())
 	}
 
-	apiExtensionClient, err := apiextensionsclient.NewForConfig(cfg)
-	if err != nil {
-		klog.Fatalf("Error creating api extension client: %s", err.Error())
-	}
-
-	err = v1alpha1.CreateCRD(apiExtensionClient)
-	if err != nil {
-		klog.Fatalf("Error creating crd: %s", err.Error())
-	}
-
 	queues := queue.NewQueues()
 	go queues.Sync(stopCh)
 

hack/install.sh

@@ -2,32 +2,33 @@
 
 set -e
 
+if [ $# -eq 0 ]; then
+  WPA_TAG=`curl -s https://api.github.com/repos/practo/k8s-worker-pod-autoscaler/releases/latest|python -c "import json;import sys;sys.stdout.write(json.load(sys.stdin)['tag_name']+'\n')"`
+else
+  WPA_TAG=$1
+fi
+
+crd='./artifacts/crd.yaml'
 serviceaccount='./artifacts/serviceaccount.yaml'
 clusterrole='./artifacts/clusterrole.yaml'
 clusterrolebinding='./artifacts/clusterrolebinding.yaml'
 new_deployment='./artifacts/deployment.yaml'
 template_deployment='./artifacts/deployment-template.yaml'
 
-if [ -z "${AWS_REGIONS}" ]; then
-    echo "AWS_REGIONS not set in environment, exiting"
-    exit 1
-fi
-if [ -z "${AWS_ACCESS_KEY_ID}" ]; then
-    echo "AWS_ACCESS_KEY_ID not set in environment, exiting"
-    exit 1
-fi
-if [ -z "${AWS_SECRET_ACCESS_KEY}" ]; then
-    echo "AWS_SECRET_ACCESS_KEY not set in environment, exiting"
-    exit 1
-fi
+echo "Creating CRD..."
+kubectl apply -f ${crd}
 
+echo "Generating Deployment Manifest..."
 export WPA_AWS_REGIONS="${AWS_REGIONS}"
 export WPA_AWS_ACCESS_KEY_ID="${AWS_ACCESS_KEY_ID}"
 export WPA_AWS_SECRET_ACCESS_KEY="${AWS_SECRET_ACCESS_KEY}"
 
+echo "Image to be used: practodev/${WPA_TAG}"
+
 cp -f $template_deployment $new_deployment
 ./hack/generate.sh ${new_deployment}
 
+echo "Applying manifests.."
 kubectl apply -f ${serviceaccount}
 kubectl apply -f ${clusterrole}
 kubectl apply -f ${clusterrolebinding}

pkg/apis/workerpodautoscaler/v1/doc.go

@@ -0,0 +1,5 @@
+// +k8s:deepcopy-gen=package
+// +groupName=k8s.practo.dev
+
+// Package v1 is the v1 version of the API.
+package v1 // import "github.com/practo/k8s-worker-pod-autoscalers/pkg/apis/workerpodautoscaler/v1"

pkg/apis/workerpodautoscaler/v1/get_default.go

@@ -0,0 +1,8 @@
+package v1
+
+func (w *WorkerPodAutoScaler) GetMaxDisruption(defaultDisruption string) *string {
+	if w.Spec.MaxDisruption == nil {
+		return &defaultDisruption
+	}
+	return w.Spec.MaxDisruption
+}