deployment

name	deployment
description	Deployment workflows, containerization, and cloud-native practices
version	1.0.0
category	DevOps
tags	["deployment","docker","k8s","cloud-run","ci-cd"]
author	ZGO Team
updated	"2026-01-24T00:00:00.000Z"

Deployment Skill

📋 Purpose

This skill defines the deployment workflows and infrastructure standards for the ZGO project. it ensures that applications are delivered consistently, securely, and with high availability across environments.

🎯 When to Use

• Building and containerizing an application
• Setting up CI/CD pipelines
• Deploying to Google Cloud Run or Kubernetes
• Managing environment-specific configurations
• performing blue-green or canary deployments

⚙️ Prerequisites

• Docker installed locally
• Access to cloud provider (Google Cloud recommended)
• Understands image registry (GCR/AR)
• Knowledge of multi-stage Docker builds

---

🏗️ Containerization

We use Multi-Stage Docker Builds to ensure small, secure production images.

1. Multi-Stage Dockerfile (Standard)

# Build Stage
FROM golang:1.21-alpine AS builder
WORKDIR /app
COPY go.mod go.sum ./
RUN go mod download
COPY . .
RUN CGO_ENABLED=0 GOOS=linux go build -o main ./cmd/server

# Production Stage
FROM alpine:latest
RUN apk --no-cache add ca-certificates tzdata
WORKDIR /root/
COPY --from=builder /app/main .
COPY --from=builder /app/config ./config
EXPOSE 8080
CMD ["./main"]

2. Best Practices

• ✅ Use .dockerignore to keep image small
• ✅ Use distroless or alpine for production to reduce attack surface
• ✅ Set CGO_ENABLED=0 for static binary
• ✅ Include non-root user in the final image

---

🛰️ Cloud Run Deployment (Recommended)

Google Cloud Run is the preferred platform for ZGO's serverless microservices.

1. Manual Deployment

# Build and push to registry
docker build -t gcr.io/[PROJECT_ID]/zgo-app:v1 .
docker push gcr.io/[PROJECT_ID]/zgo-app:v1

# Deploy to Cloud Run
gcloud run deploy zgo-app \
  --image gcr.io/[PROJECT_ID]/zgo-app:v1 \
  --region us-central1 \
  --allow-unauthenticated

2. AI-Assisted Deployment (MCP)

If the cloudrun MCP server is enabled, use it to deploy directly from the context.

---

🔄 CI/CD Workflow

We follow a GitOps approach using GitHub Actions.

Pipeline Stages

1. Lint: Run golangci-lint
2. Test: Execute unit and integration tests
3. Build: Create Docker image
4. Scan: Vulnerability scanning (Trivy/Snyk)
5. Deploy: Auto-deploy to Staging; Manual trigger to Production

---

⚙️ Environment Management

Configurations must be injected via Environment Variables for 12-factor compliance.

Key	Example (Dev)	Example (Prod)	Description
APP_ENV	development	production	Environment toggle
DB_DSN	localhost:5432	db.production.local	Database connection
LOG_LEVEL	debug	info	Log verbosity
SECRET_KEY	dev-secret	KMS-stored-secret	Security token

---

🩺 Health Checks & Monitoring

1. Standard Endpoints

• /healthz: Liveness check (checks process status)
• /readyz: Readiness check (checks DB/Redis connectivity)

2. Observability

• Metrics: Exported via Prometheus (default port: 9090)
• Traces: Exported to Cloud Trace / Jaeger
• Logs: Structured JSON logs sent to stdout

---

✅ Verification Checklist

• Dockerfile uses multi-stage builds.
• No secrets are hardcoded in the Dockerfile.
• .dockerignore excludes large/sensitive files (vendor, .env, tmp).
• Liveness/Readiness probes are implemented.
• Image size is under 100MB (target: ~30-50MB).
• CI/CD pipeline passes all stages.

---

🔧 Automation Scripts

Available in scripts/:

• deploy-cloudrun.sh - Helper for GCR deployments.
• verify-image.sh - Checks image compliance & size.

---

📚 Complete Examples

• GitHub Actions Workflow YAML
• Full Production Dockerfile
• Health Check Implementation

---

🔗 Related Skills

• coding-standards: For log and config standards.
• api-development: For health check endpoint standards.

---

Version: 1.0.0
Last Updated: 2026-01-24
Maintainer: ZGO Team