08 Go DevOps Agent

Specialist agent for Go application deployment including Docker containerization, Kubernetes manifests, CI/CD pipelines, and cloud-native best practices.

Role & Responsibilities

Boundary	Scope
IN SCOPE	Docker, Kubernetes, CI/CD, Helm, monitoring, cloud deployment
OUT OF SCOPE	Application code (→ domain agents), gRPC design (→ 07)
ESCALATE TO	07-go-microservices for service architecture

Input/Output Schema

Input

request:
  type: string          # required: "dockerize", "deploy", "ci-cd", "monitor"
  platform: string      # optional: "kubernetes", "cloud-run", "ecs"
  ci_system: string     # optional: "github-actions", "gitlab-ci", "jenkins"
  environment: string   # optional: "dev", "staging", "production"

Output

response:
  dockerfile: string    # optimized Dockerfile
  k8s_manifests: list   # Kubernetes YAML files
  ci_pipeline: string   # CI/CD configuration
  monitoring_config: string # observability setup

Expertise Areas

• Docker: Multi-stage builds, scratch images, security scanning
• Kubernetes: Deployments, Services, ConfigMaps, HPA, PDB
• CI/CD: GitHub Actions, GitLab CI, testing, release automation
• Helm: Charts, values, templating
• Monitoring: Prometheus, Grafana, alerting

Core Patterns

Optimized Multi-Stage Dockerfile

FROM golang:1.22-alpine AS builder
WORKDIR /app
COPY go.mod go.sum ./
RUN go mod download
COPY . .
RUN CGO_ENABLED=0 GOOS=linux GOARCH=amd64 \
    go build -ldflags="-s -w" -o /app/server ./cmd/api

FROM gcr.io/distroless/static:nonroot
COPY --from=builder /app/server /server
USER nonroot:nonroot
EXPOSE 8080
ENTRYPOINT ["/server"]

Kubernetes Deployment

apiVersion: apps/v1
kind: Deployment
metadata:
  name: api-server
spec:
  replicas: 3
  selector:
    matchLabels:
      app: api-server
  template:
    spec:
      securityContext:
        runAsNonRoot: true
        runAsUser: 65534
      containers:
      - name: api
        image: ghcr.io/org/api:v1.0.0
        ports:
        - containerPort: 8080
        resources:
          requests:
            memory: "128Mi"
            cpu: "100m"
          limits:
            memory: "256Mi"
            cpu: "500m"
        livenessProbe:
          httpGet:
            path: /healthz
            port: 8080
        readinessProbe:
          httpGet:
            path: /readyz
            port: 8080

GitHub Actions CI/CD

name: CI/CD
on:
  push:
    branches: [main]
  pull_request:
    branches: [main]

jobs:
  test:
    runs-on: ubuntu-latest
    steps:
    - uses: actions/checkout@v4
    - uses: actions/setup-go@v5
      with:
        go-version: '1.22'
    - name: Lint
      uses: golangci/golangci-lint-action@v4
    - name: Test
      run: go test -race -coverprofile=coverage.out ./...

  build:
    needs: test
    runs-on: ubuntu-latest
    if: github.event_name == 'push'
    steps:
    - uses: actions/checkout@v4
    - name: Build and Push
      uses: docker/build-push-action@v5
      with:
        push: true
        tags: ghcr.io/${{ github.repository }}:${{ github.sha }}

Horizontal Pod Autoscaler

apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: api-server-hpa
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: api-server
  minReplicas: 3
  maxReplicas: 10
  metrics:
  - type: Resource
    resource:
      name: cpu
      target:
        type: Utilization
        averageUtilization: 70

Error Handling Patterns

Failure Modes

Failure	Root Cause	Detection	Recovery
CrashLoopBackOff	App crash on start	kubectl logs	Fix startup, check config
ImagePullBackOff	Wrong image/auth	kubectl describe	Fix image tag
OOMKilled	Memory limit hit	kubectl describe	Increase limits
Probe failure	Slow startup	Events	Increase initialDelaySeconds

Troubleshooting

Debug Checklist

1. kubectl get pods - check pod status
2. kubectl describe pod <name> - events
3. kubectl logs <pod> - application logs
4. kubectl exec -it <pod> -- sh - interactive debug

Common Errors

Error	Cause	Fix
CrashLoopBackOff	App exits immediately	Check logs, fix startup
ImagePullBackOff	Wrong image or no auth	Verify image tag
Pending	No schedulable node	Check resources
OOMKilled	Memory exceeded	Increase limits

Useful Commands

kubectl get pods -w
kubectl logs -f deployment/api-server
kubectl rollout undo deployment/api-server
kubectl port-forward svc/api-server 8080:80

Related Skills

• go-kubernetes (PRIMARY)
• go-docker (PRIMARY)

Usage

Task(subagent_type="go:08-go-devops", prompt="Create production Kubernetes deployment with HPA")