Deploy Bookshelf to Container Engine/Kubernetes

This optional tutorial will walk you through how to deploy the Bookshelf sample application to Google Container Engine. This tutorial is also applicable to Kubernetes outside of Container Engine, but may require additional steps for external load balancing.

Pre-requisites

Create a project in the Google Cloud Platform Console.
Enable billing for your project.
Enable APIs for your project. The provided link will enable all necessary APIs, but if you wish to do so manually you will need Datastore, Pub/Sub, Storage, and Logging.

Install the Google Cloud SDK

 $ curl https://sdk.cloud.google.com | bash 
 $ gcloud init

Install Docker.

Create a cluster

Create a cluster for the bookshelf application:

gcloud container clusters create bookshelf \
    --scopes "cloud-platform" \
    --num-nodes 2
gcloud container clusters get-credentials bookshelf

The scopes specified in the --scopes argument allows nodes in the cluster to access Google Cloud Platform APIs, such as the Cloud Datastore API.

Alternatively, you can use make:

make create-cluster

Create a Cloud Storage bucket

The bookshelf application uses Google Cloud Storage to store image files. Create a bucket for your project:

gsutil mb gs://<your-project-id>
gsutil defacl set public-read gs://<your-project-id>

Alternatively, you can use make:

make create-bucket

Update config.py

Modify config.py and enter your Cloud Project ID into the PROJECT_ID and CLOUD_STORAGE_BUCKET field. The remaining configuration values are only needed if you wish to use a different database or if you wish to enable log-in via oauth2, which requires a domain name.

Build the bookshelf container

Before the application can be deployed to Container Engine, you will need build and push the image to Google Container Registry.

docker build -t gcr.io/your-project-id/bookshelf .
gcloud docker push gcr.io/your-project-id/bookshelf

Alternativel,y you can use make:

make push

Deploy the bookshelf frontend

The bookshelf app has two distinct "tiers". The frontend serves a web interface to create and manage books, while the worker handles fetching book information from the Google Books API.

Update bookshelf-frontend.yaml with your Project ID or use make template. This file contains the Kubernetes resource definitions to deploy the frontend. You can use kubectl to create these resources on the cluster:

kubectl create -f bookshelf-frontend.yaml

Alternatively, you can use make:

make deploy-frontend

Once the resources are created, there should be 3 bookshelf-frontend pods on the cluster. To see the pods and ensure that they are running:

kubectl get pods

If the pods are not ready or if you see restarts, you can get the logs for a particular pod to figure out the issue:

kubectl logs pod-id

Once the pods are ready, you can get the public IP address of the load balancer:

kubectl get services bookshelf-frontend

You can then browse to the public IP address in your browser to see the bookshelf application.

Deploy worker

Update bookshelf-worker.yaml with your Project ID or use make template. This file contains the Kubernetes resource definitions to deploy the worker. The worker doesn't need to serve web traffic or expose any ports, so it has significantly less configuration than the frontend. You can use kubectl to create these resources on the cluster:

kubectl create -f bookshelf-worker.yaml

Alternatively, you can use make:

make deploy-backend

Once again, use kubectl get pods to check the status of the worker pods. Once the worker pods are up and running, you should be able to create books on the frontend and the workers will handle updating book information in the background.