Using the Cloud Spanner Emulator in CI/CD pipelines

In previous posts, we now have seen how one can provision and handle Cloud Spanner in manufacturing environments and how one can use the Cloud Spanner Emulator in your improvement workflow with a pattern Node.js app referred to as OmegaTrade. We’ve coated:

1. The Cloud Spanner emulator’s numerous deployment fashions,

2. Operating the emulator regionally with OmegaTrade, and 

3. Operating the emulator remotely with OmegaTrade

On this submit, we’ll cowl how one can use the Cloud Spanner emulator in your Steady Integration and Steady Supply/Deployment (CI/CD) pipelines, with pattern steps protecting the next instruments: 

1. Cloud Construct

2. Jenkins

3. GitHub Actions

4. CircleCI

As a refresher, OmegaTrade is a inventory ticker visualization software that shops inventory costs in Cloud Spanner and renders visualizations utilizing Google Charts. It consists of a frontend service that renders the visualizations and a backend service that writes to and reads from the Cloud Spanner occasion. 

For the sake of simplicity, we concentrate on the backend service of the OmegaTrade software on this weblog submit and reveal how one can use the Cloud Spanner emulator in CI/CD pipelines for this service. 

For those who’d desire to deploy the OmegaTrade backend manually, you’ll be able to study all about how to try this by studying this weblog submit and specializing in the part Guide deployment steps.

CI/CD Method for deploying OmegaTrade

The steps for all the CI/CD instruments we focus on are related. The code is saved in a public GitHub repository, Docker recordsdata are used for creating the appliance docker picture, Google Container Registry is used because the docker picture repository and gcloud instructions are used for software deployment to Cloud Run.

Please observe that all the CI/CD instruments require entry to a GCP Service Account (SA) in addition to integration together with your GitHub repository. 

The principle department retains the appliance code to be deployed on the dev surroundings the place we will likely be deploying the Cloud Spanner emulator. Since it’s utilizing the emulator, we’re calling this a dev surroundings.

The prod department retains the appliance code to be deployed on the prod surroundings the place we will likely be deploying an precise Cloud Spanner occasion. 

Now we have two dockerfiles, one for dev (dockerfile.native.emulator) the place we will likely be deploying the Cloud Spanner emulator, and one other for prod (dockerfile.prod). dockerfile.native.emulator comprises the Cloud Spanner emulator in addition to the appliance layers (for testing functions) whereas dockerfile.prod solely comprises the appliance code layer.

Cloud Construct

To arrange CI/CD with Cloud Construct, please make sure the Cloud Construct API is enabled out of your Cloud Console as a prerequisite. Step one is to attach our GitHub repository with Cloud Construct for automated deployment of our software over Cloud Run. 

From the Join Repository possibility within the Cloud Construct Triggers display screen, choose the supply as GitHub (Cloud Construct GitHub App), authenticate, choose GitHub Account and Repository.

Within the backend, Cloud Construct will set up the Cloud Construct app in your GitHub repository. You’ll find Cloud Construct in Repo Settings ➞ Integrations. This app will monitor your repository and set off pipeline processing upon any commit or push to whichever department you point out within the set off. Cloud Construct permits folder-specific triggers.

For this weblog, we’re going to create 2 triggers, one devoted to dev deployment and one other to prod. The Dev deployment will likely be speaking with the Spanner emulator whereas Prod one will likely be speaking with an precise Cloud Spanner occasion.

Create Google Cloud Construct Triggers

We’re utilizing the Cloud Construct.yaml file (construct config file) for the deployment. A construct config file defines the fields which can be wanted for Cloud Construct to carry out your duties. We will create both a .yml or .json file for Cloud Construct the place we write directions for our CI/CD pipeline.

The Cloud Construct config is only parameterized which suggests we’re utilizing the identical construct config file for our frontend in addition to backend deployment and offering values whereas creating triggers within the GCP UI.

The substitution_option: ‘ALLOW_LOOSE‘ permits Cloud Construct triggers to run regardless of any lacking substitution variable. We’re utilizing this selection as a result of we’d like some additional values for backend deployment. On this case, Cloud Construct received’t return any errors.

To arrange the backend set off, comply with the under steps:

• From Cloud Construct ➞ Create Set off, enter a singular title, description for the set off

• Choose Occasion as Push to a department

• Select GitHub repo in Supply 

• Enter ^fundamental$ beneath Department

• Enter backend/** within the Included recordsdata filter (this selection triggers a backend set off when there may be any change(s) within the folder)

• Enter Cloud Construct.yaml in Construct Configuration.