Canary-deploy an application to a target

Canary-deploy an application to a target

This quickstart shows you how to use Cloud Deploy to deliver a sample application image in a canary deployment to either Google Kubernetes Engine or to Cloud Run. (You can also run a canary deployment to GKE Enterprise, but only GKE and Cloud Run are shown in this quickstart.)

A canary deployment splits traffic between an already-deployed version of the application and the new version. Cloud Run apportions traffic based on the percentages you configure in the delivery pipeline. GKE deploys the new version to a proportion of pods. This quickstart deploys to 50% first, then to 100%.

In this quickstart, there is only one target, (prod). So we create only one GKE cluster or one Cloud Run service to run your application.

In this quickstart, you'll do the following:

1. Create one GKE cluster or define one Cloud Run service.

   You can canary deploy to GKE Enterprise clusters too, but this quickstart uses GKE and Cloud Run only.

2. Create a Skaffold configuration and a Kubernetes manifest to specify the (pre-built) container image to deploy.

3. Define your Cloud Deploy delivery pipeline and deployment target.

4. Invoke your delivery pipeline by creating a release, which automatically deploys to one target.

   This first release skips the canary phase.

5. View the delivery pipeline and release in the Google Cloud console.

6. Create a second release, this time executing the canary stage to deploy the application to 50%.

7. Advance the release to deploy to 100%.

Before you begin

1. Sign in to your Google Cloud account. If you're new to Google Cloud, create an account to evaluate how our products perform in real-world scenarios. New customers also get $300 in free credits to run, test, and deploy workloads.

2. In the Google Cloud console, on the project selector page, select or create a Google Cloud project.

   Note: If you don't plan to keep the resources that you create in this procedure, create a project instead of selecting an existing project. After you finish these steps, you can delete the project, removing all resources associated with the project.

   Go to project selector

3. Make sure that billing is enabled for your Google Cloud project.

4. Enable the Cloud Deploy, Cloud Build, GKE, Cloud Run, and Cloud Storage APIs.

   Enable the APIs

5. Install the Google Cloud CLI.

6. To initialize the gcloud CLI, run the following command:

   gcloud init

   Note: You can run the gcloud CLI in the Google Cloud console without installing the Google Cloud CLI. To run the gcloud CLI in the Google Cloud console, use Cloud Shell.

7. In the Google Cloud console, on the project selector page, select or create a Google Cloud project.

   Note: If you don't plan to keep the resources that you create in this procedure, create a project instead of selecting an existing project. After you finish these steps, you can delete the project, removing all resources associated with the project.

   Go to project selector

8. Make sure that billing is enabled for your Google Cloud project.

9. Enable the Cloud Deploy, Cloud Build, GKE, Cloud Run, and Cloud Storage APIs.

   Enable the APIs

10. Install the Google Cloud CLI.

11. To initialize the gcloud CLI, run the following command:

    gcloud init

    Note: You can run the gcloud CLI in the Google Cloud console without installing the Google Cloud CLI. To run the gcloud CLI in the Google Cloud console, use Cloud Shell.

If you already have the CLI installed, make sure you're running the latest version:

gcloud components update

12. Make sure the default Compute Engine service account has sufficient permissions.

    The service account might already have the necessary permissions. These steps are included for projects that disable automatic role grants for default service accounts.

    1. First add the clouddeploy.jobRunner role:

       gcloud projects add-iam-policy-binding PROJECT_ID \
           --member=serviceAccount:$(gcloud projects describe PROJECT_ID \
           --format="value(projectNumber)")[e-mail] \
           --role="roles/clouddeploy.jobRunner"
       

    2. Add the developer role for your specific runtime.
    • For GKE, and GKE with Gateway API:

      gcloud projects add-iam-policy-binding PROJECT_ID \
          --member=serviceAccount:$(gcloud projects describe PROJECT_ID \
          --format="value(projectNumber)")[e-mail] \
          --role="roles/container.developer"
      

    • For Cloud Run:

      gcloud projects add-iam-policy-binding PROJECT_ID \
          --member=serviceAccount:$(gcloud projects describe PROJECT_ID \
          --format="value(projectNumber)")[e-mail] \
          --role="roles/run.developer"
      

    3. Add the iam.serviceAccountUser role, which includes the actAspermission to deploy to the runtime:

       gcloud iam service-accounts add-iam-policy-binding $(gcloud projects describe PROJECT_ID \
           --format="value(projectNumber)")[e-mail] \
           --member=serviceAccount:$(gcloud projects describe PROJECT_ID \
           --format="value(projectNumber)")[e-mail] \
           --role="roles/iam.serviceAccountUser" \
           --project=PROJECT_ID
       

Create your runtime environment

GKE

Create one GKE Autopilot cluster:

 gcloud container clusters create-auto canary-quickstart-cluster \
                  --project=PROJECT_ID \
                  --region=us-central1

GKE + Gateway API

1. Create one GKE cluster, with recommended settings to support using with Istio:

   gcloud container clusters create canary-quickstart-cluster \
          --machine-type=n1-standard-1 \
          --num-nodes 4 \
          --region=us-central1 \
          --project=PROJECT_ID
   

2. Get the cluster credentials:

   gcloud container clusters get-credentials canary-quickstart-cluster \
          --project=PROJECT_ID \
          --region=us-central1
   

3. Install the Kubernetes Gateway API CRDs if not already present on the cluster.

   kubectl apply -f https://github.com/kubernetes-sigs/gateway-api/releases/download/v0.6.2/standard-install.yaml
   

4. Enable Istio's Gateway controller implementation by installing Istio.

   curl -L https://istio.io/downloadIstio | ISTIO_VERSION=1.17.2 sh - \
   && ./istio-1.17.2/bin/istioctl install --set profile=minimal -y
   

Cloud Run

If you're using Cloud Run, you can skip this command, you don't need to do anything here.

Prepare your Skaffold configuration and application manifest

Cloud Deploy uses Skaffold to provide the details for what to deploy and how to deploy it properly to your target.

In this quickstart, you create a skaffold.yaml file, which identifies the Kubernetes manifest or Cloud Run service configuration to be deployed.

1. Open a terminal window.

2. Create a new directory and navigate into it.

   GKE

   mkdir deploy-canary-quickstart-gke
   cd deploy-canary-quickstart-gke
   

   GKE + Gateway API

   mkdir deploy-canary-quickstart-gke-gatewayapi
   cd deploy-canary-quickstart-gke-gatewayapi
   

   Cloud Run

   mkdir deploy-canary-quickstart-run
   cd deploy-canary-quickstart-run
   

3. Create a file named skaffold.yaml with the following contents:

   GKE

   apiVersion: skaffold/v4beta7
   kind: Config
   manifests:
     rawYaml:
     - kubernetes.yaml
   deploy:
     kubectl: {}
   

   GKE + Gateway API

   apiVersion: skaffold/v4beta7
   kind: Config
   manifests:
     rawYaml:
     - kubernetes.yaml
   deploy:
     kubectl: {}
   

   Cloud Run

   apiVersion: skaffold/v4beta7
   kind: Config
   manifests:
     rawYaml:
     - run.yaml
   deploy:
     cloudrun: {}
   

   This file is a minimal Skaffold config, identifying your manifest. For this quickstart, you create the file. But you can also have Cloud Deploy create one for you, for simple, non-production applications.

   See the skaffold.yaml reference for more information about this file.

4. Create your application manifest.

   GKE

   Create a file named kubernetes.yaml, in the deploy-canary-quickstart-gke directory, with the following contents:

   apiVersion: apps/v1
   kind: Deployment
   metadata:
     name: my-deployment
     labels:
       app: my-app
     namespace: default
   spec:
     replicas: 1
     selector:
       matchLabels:
         app: my-app
     template:
       metadata:
         labels:
           app: my-app
       spec:
         containers:
         - name: nginx
           image: my-app-image
   ---
   apiVersion: v1
   kind: Service
   metadata:
     name: my-service
     namespace: default
   spec:
     selector:
       app: my-app
     ports:
       - protocol: TCP
         port: 80
   

   This file is a simple Kubernetes manifest, which is applied to the cluster to deploy the application.

   GKE + Gateway API

   Create a file named kubernetes.yaml, in the deploy-canary-quickstart-gke-gatewayapi directory, with the following contents:

   kind: Gateway
   apiVersion: gateway.networking.k8s.io/v1beta1
   metadata:
     name: my-gateway
     annotations:
       networking.istio.io/service-type: "ClusterIP"
   spec:
     gatewayClassName: istio
     listeners:
     - name: default
       hostname: "*.example.com"
       port: 80
       protocol: HTTP
       allowedRoutes:
         namespaces:
           from: All
   ---
   kind: HTTPRoute
   apiVersion: gateway.networking.k8s.io/v1beta1
   metadata:
     name: my-httproute
   spec:
     parentRefs:
     - kind: Gateway
       name: my-gateway
     hostnames:
     - "test.example.com"
     rules:
     - backendRefs:
       - name: my-service
         port: 80
   ---
   apiVersion: v1
   kind: Service
   metadata:
     name: my-service
   spec:
     selector:
       app: my-app
     ports:
     - name: tcp-port
       protocol: TCP
       port: 80
       targetPort: 8080
   ---
   apiVersion: apps/v1
   kind: Deployment
   metadata:
     name: my-deployment
     labels:
       app: my-app
   spec:
     replicas: 1
     selector:
       matchLabels:
         app: my-app
     template:
       metadata:
         labels:
           app: my-app
       spec:
         containers:
         - name: nginx
           image: my-app-image
   

   This file is a Kubernetes manifest, which is applied to the cluster to deploy the application. This manifest includes the Service and Deployment resources required for canary deployment, plus an HTTPRoute and the Gateway resource needed for using Gateway API.

   Cloud Run

   Create a file named run.yaml, in the deploy-canary-quickstart-run directory, with the following contents:

   apiVersion: serving.knative.dev/v1
   kind: Service
   metadata:
     name: my-canary-run-service
   spec:
     template:
       spec:
         containers:
         - image: my-app-image
   

   This file is a Cloud Run service configuration file, which is applied at deploy time to create your service in Cloud Run.

Note: If you want to use different manifests per target, read this article about managing manifests to find out more about using Skaffold profiles.

Create your delivery pipeline and targets

You can define your delivery pipeline and targets in one file or in separate files. In this quickstart, we create one file for our pipeline and our single target:

GKE

Create a file named clouddeploy.yaml, in the deploy-canary-quickstart-gke directory, with the following contents:

apiVersion: deploy.cloud.google.com/v1
kind: DeliveryPipeline
metadata:
  name: my-canary-demo-app-1
description: main application pipeline
serialPipeline:
  stages:
  - targetId: prod
    profiles: []
    strategy:
      canary:
        runtimeConfig:
          kubernetes:
            serviceNetworking:
              service: "my-service"
              deployment: "my-deployment"
        canaryDeployment:
          percentages: [50]
          verify: false
---
apiVersion: deploy.cloud.google.com/v1
kind: Target
metadata:
  name: prod
description: prod GKE cluster
gke:
 cluster: projects/PROJECT_ID/locations/us-central1/clusters/canary-quickstart-cluster

GKE + Gateway API

Create a file named clouddeploy.yaml, in the deploy-canary-quickstart-gke-gatewayapi directory, with the following contents:

apiVersion: deploy.cloud.google.com/v1
kind: DeliveryPipeline
metadata:
  name: my-canary-demo-app-1
description: main application pipeline
serialPipeline:
  stages:
  - targetId: prod
    profiles: []
    strategy:
      canary:
        runtimeConfig:
          kubernetes:
            gatewayServiceMesh:
              httpRoute: "my-httproute"
              service: "my-service"
              deployment: "my-deployment"
        canaryDeployment:
          percentages: [50]
          verify: false
---
apiVersion: deploy.cloud.google.com/v1
kind: Target
metadata:
  name: prod
description: prod GKE cluster
gke:
 cluster: projects/PROJECT_ID/locations/us-central1/clusters/canary-quickstart-cluster

Cloud Run

Create a file named clouddeploy.yaml, in the deploy-canary-quickstart-run directory, with the following contents:

apiVersion: deploy.cloud.google.com/v1
kind: DeliveryPipeline
metadata:
  name: my-canary-demo-app-1
description: main application pipeline
serialPipeline:
  stages:
  - targetId: prod
    profiles: []
    strategy:
      canary:
        runtimeConfig:
          cloudRun:
            automaticTrafficControl: true
        canaryDeployment:
          percentages: [50]
          verify: false
---
apiVersion: deploy.cloud.google.com/v1
kind: Target
metadata:
  name: prod
description: prod Run Service
run:
  location: projects/PROJECT_ID/locations/us-central1

1. Register your pipeline and targets with the Cloud Deploy service:

   gcloud deploy apply --file=clouddeploy.yaml --region=us-central1 --project=PROJECT_ID
   

   You now have a pipeline, with one target configured for a canary deployment strategy.

2. Confirm your pipeline and targets:

   In the Google Cloud console, navigate to the Cloud Deploy Delivery pipelines page to view of list of your available delivery pipelines.

   Open the Delivery pipelines page

   The delivery pipeline you just created is shown, and the one target you configured is listed in the Targets column.

   

Create a release

A release is the central Cloud Deploy resource representing the changes being deployed. The delivery pipeline defines the lifecycle of that release. See Cloud Deploy service architecture for details about that lifecycle.

To create a release resource that represents the container image to deploy, run the following command from the deploy-canary-quickstart-gke, deploy-canary-quickstart-gke-gatewayapi, or deploy-canary-quickstart-run directory:

GKE

 gcloud deploy releases create test-release-001 \
   --project=PROJECT_ID \
   --region=us-central1 \
   --delivery-pipeline=my-canary-demo-app-1 \
   --images=my-app-image=gcr.io/google-containers/nginx@sha256:f49a843c290594dcf4d193535d1f4ba8af7d56cea2cf79d1e9554f077f1e7aaa

GKE + Gateway API

 gcloud deploy releases create test-release-001 \
   --project=PROJECT_ID \
   --region=us-central1 \
   --delivery-pipeline=my-canary-demo-app-1 \
   --images=my-app-image=gcr.io/google-containers/nginx@sha256:f49a843c290594dcf4d193535d1f4ba8af7d56cea2cf79d1e9554f077f1e7aaa

Cloud Run

 gcloud deploy releases create test-release-001 \
   --project=PROJECT_ID \
   --region=us-central1 \
   --delivery-pipeline=my-canary-demo-app-1 \
   --images=my-app-image=us-docker.pkg.dev/cloudrun/container/hello@sha256:6063adf8f687702b4065151acddba6781c47bc602167eb9f3bec8aebc9ce95cc

When you create a release, Cloud Deploy automatically creates a rollout resource too, to immediately deploy to your one target, prod.

We skip to the stable phase

With this first release, we skip the canary phase, and deploy to 100% (stable phase). This is because the application hasn't been deployed previously, so there's no way to calculate 50% of pods (for GKE) or how traffic is split for the service (for Cloud Run). The pods (GKE) or revisions (Cloud Run) don't exist yet.

Having skipped the canary phase, we're now ready to start the stable phase, which takes traffic to 100%. After that, we'll create another release, and that will execute the canary.

In a real-world situation, you will usually execute a canary deployment where your application is already running, so this phase skipping will be rare.

View the release in Google Cloud console

Now that you've created the first release, the rollout is created, and you can view the release and the rollout in Google Cloud console. You can also view the pipeline visualization, which shows the current status of the release.

1. In the Google Cloud console, navigate to the Cloud Deploy Delivery pipelines page to view your my-canary-demo-app-1 delivery pipeline.

   Open the Delivery pipelines page

2. Click the name of your delivery pipeline "my-canary-demo-app-1".

   The pipeline visualization shows the app's deployment status. Because there's only one stage in the pipeline, the visualization shows only one node.

   

   And your release is listed on the Releases tab under Delivery pipeline details.

3. Click the release name, test-release-001.

   Your rollouts appear under Rollouts. You can click a rollout to view its details, including the deployment log.

   

   Notice that the rollout status is "Pending advance," and the target shown in the pipeline visualization has a link to "Advance to stable."

Advance the rollout phase

After the first release, the canary phase was skipped, and the rollout is waiting to start the "stable" phase, which deploys the application to 100%:

1. In the pipeline visualization, click Advance to stable.

2. When prompted, click Advance to confirm.

After a few minutes, the rollout is now in the "stable" phase, and the application is deployed to 100%.

Execute the canary deployment

Because the first release skipped the canary phase, we'll now create another release, which this time does execute a canary deployment.

1. To create a new release, run the following command from the deploy-canary-quickstart-gke, deploy-canary-quickstart-gke-gatewayapi or deploy-canary-quickstart-run directory:

   GKE

   gcloud deploy releases create test-release-002 \
     --project=PROJECT_ID \
     --region=us-central1 \
     --delivery-pipeline=my-canary-demo-app-1 \
     --images=my-app-image=gcr.io/google-containers/nginx@sha256:f49a843c290594dcf4d193535d1f4ba8af7d56cea2cf79d1e9554f077f1e7aaa
   

   GKE + Gateway API

   gcloud deploy releases create test-release-002 \
     --project=PROJECT_ID \
     --region=us-central1 \
     --delivery-pipeline=my-canary-demo-app-1 \
     --images=my-app-image=gcr.io/google-containers/nginx@sha256:f49a843c290594dcf4d193535d1f4ba8af7d56cea2cf79d1e9554f077f1e7aaa
   

   Cloud Run

   gcloud deploy releases create test-release-002 \
     --project=PROJECT_ID \
     --region=us-central1 \
     --delivery-pipeline=my-canary-demo-app-1 \
     --images=my-app-image=us-docker.pkg.dev/cloudrun/container/hello@sha256:6063adf8f687702b4065151acddba6781c47bc602167eb9f3bec8aebc9ce95cc
   

   After a few minutes, a rollout is created, and this time the canary stage is executed:

   

   When the first rollout phase finishes, the rollout is now in the canary phase:

   

   This means that the application is now deployed to 50%. For serviceNetworking-based GKE, it's deployed to half of your pods. For Gateway API-based GKE and Cloud Run traffic is allocated to 50%.

2. Click Advance Rollout, then click Advance when prompted.

   This advances the rollout to the "stable" phase, deploying the application to 100%.

   

Clean up

To avoid incurring charges to your Google Cloud account for the resources used on this page, follow these steps.

1. Delete the canary-quickstart-cluster cluster (GKE only):

   gcloud container clusters delete canary-quickstart-cluster --region=us-central1 --project=PROJECT_ID
   

2. Delete the my-canary-run-service service (Cloud Run only):

   gcloud run services delete my-canary-run-service --region=us-central1 --project=PROJECT_ID
   

3. Delete the delivery pipeline, target, and all release and rollout resources:

   gcloud deploy delete --file=clouddeploy.yaml --force --region=us-central1 --project=PROJECT_ID
   

4. Delete the Cloud Storage buckets that Cloud Deploy created.

   One ends with _clouddeploy, and the other is [region].deploy-artifacts.[project].appspot.com.

   Open the Cloud Storage browser page

That's it, you completed this quickstart!

What's next

• Learn more about Cloud Deploy.

• Learn the basics of deploying applications.

• Try out the Cloud Deploy walkthrough.

• Learn how to manage your manifests.

• Learn how to combine Google Cloud CI/CD tools to develop and deliver software effectively to GKE.