gcp-deployment

name	gcp-deployment
description	Deep knowledge about deploying applications to Google Cloud Platform (Cloud Run, App Engine, Compute Engine).
type	transform
tier	library
domain	deploy
trigger	when deploying to GCP

Context

Deploying {{project_name}} ({{project_type}}) to Google Cloud Platform (GCP). You will follow a strict 6-phase Deployment Lifecycle Contract.

Instructions

Execute the following phases in order:

Phase 1: Authentication

1. Ensure the GOOGLE_APPLICATION_CREDENTIALS environment variable is set to a valid service account key JSON file, or configure Workload Identity Federation (if inside CI/CD).
2. Authenticate the gcloud CLI: gcloud auth activate-service-account --key-file=$GOOGLE_APPLICATION_CREDENTIALS.
3. Set the default project: gcloud config set project <PROJECT_ID>.

Phase 2: Build

1. Prepare the artifacts for deployment.
2. If Cloud Run or GKE: build your Docker image (e.g., docker build -t gcr.io/<PROJECT_ID>/<APP_NAME> . or gcloud builds submit --tag gcr.io/<PROJECT_ID>/<APP_NAME>).
3. If App Engine: prepare app.yaml.
4. If Static Site (Cloud Storage): run npm run build.

Phase 3: Install / Provisioning

1. Ensure the target GCP resources (Cloud Run services, Buckets, Compute Instances) exist and you have permissions to write to them.
2. If Cloud Storage: gsutil ls gs://<bucket-name>.
3. Enable necessary APIs (e.g., gcloud services enable run.googleapis.com if using Cloud Run).

Phase 4: Deploy

1. Ship the artifact to GCP.
2. Cloud Run: gcloud run deploy <SERVICE_NAME> --image gcr.io/<PROJECT_ID>/<APP_NAME> --region <REGION> --platform managed.
3. App Engine: gcloud app deploy.
4. Cloud Storage: gsutil rsync -R dist/ gs://<bucket-name>.

Phase 5: Checking

1. Verify the deployment was successful.
2. Cloud Run: Check the returned service URL via curl and inspect gcloud logging read "resource.type=cloud_run_revision AND resource.labels.service_name=<SERVICE_NAME>" if it fails.
3. App Engine: Check the target URL.
4. Ensure the service returns HTTP 200.

Phase 6: Update / Rollback

1. If Phase 5 fails, immediately initiate a rollback.
2. Cloud Run: Route 100% of traffic to the previous revision (gcloud run services update-traffic <SERVICE_NAME> --to-revisions=<PREVIOUS_REVISION_NAME>=100 --region <REGION>).
3. App Engine: Roll back traffic to an older version via gcloud app services set-traffic.
4. Note the failure in the progress log.

Validation

• GCP authentication succeeds.
• Build succeeds.
• Artifacts are deployed.
• Health check (curl) returns 200 OK.