GitLab Runner Autoscaling on Kubernetes

# ตัวอย่าง HPA configuration
apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: gitlab-runner
  namespace: gitlab-runners
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: gitlab-runner-gitlab-runner
  minReplicas: 0
  maxReplicas: 10
  
  # Scale based on GitLab queue depth
  behavior:
    scaleUp:
      stabilizationWindowSeconds: 30
      policies:
        - type: Percent
          value: 100
          periodSeconds: 30
        - type: Pods
          value: 2
          periodSeconds: 30
      selectPolicy: Max
    scaleDown:
      stabilizationWindowSeconds: 300
      policies:
        - type: Percent
          value: 100
          periodSeconds: 15
        - type: Pods
          value: 1
          periodSeconds: 60
      selectPolicy: Min

# Optimized Pod Template
apiVersion: v1
kind: Pod
metadata:
  name: gitlab-runner-pod-template
  namespace: gitlab-runners
  labels:
    app: gitlab-runner
spec:
  # Security Context
  securityContext:
    runAsNonRoot: true
    runAsUser: 1000
    fsGroup: 1000
  
  # Prefer low-latency nodes
  affinity:
    nodeAffinity:
      preferredDuringSchedulingIgnoredDuringExecution:
      - weight: 100
        preference:
          matchExpressions:
          - key: node-type
            operator: In
            values: ["compute-optimized"]
  
  # Containers
  containers:
  - name: gitlab-runner
    image: gitlab/gitlab-runner:latest
    securityContext:
      allowPrivilegeEscalation: false
      readOnlyRootFilesystem: true
      capabilities:
        drop:
        - ALL
    
    # Resource limits
    resources:
      requests:
        memory: "256Mi"
        cpu: "100m"
      limits:
        memory: "2Gi"
        cpu: "1000m"
    
    # Environment
    env:
    - name:-ci-runner
      value: "true"
    
    # Volume mounts
    volumeMounts:
    - name: cache
      mountPath: /cache
    - name: config
      mountPath: /etc/gitlab-runner
      readOnly: true
  
  # Volumes
  volumes:
  - name: cache
    emptyDir: {}
  - name: config
    configMap:
      name: gitlab-runner-config

# Advanced HPA with custom metrics
apiVersion: autoscaling/v2
kind: HorizontalPodAutoscaler
metadata:
  name: gitlab-runner-advanced
  namespace: gitlab-runners
spec:
  scaleTargetRef:
    apiVersion: apps/v1
    kind: Deployment
    name: gitlab-runner-gitlab-runner
  minReplicas: 0
  maxReplicas: 20
  
  behavior:
    # Fast scale-up for burst jobs
    scaleUp:
      stabilizationWindowSeconds: 60
      policies:
      - type: Pods
        value: 4
        periodSeconds: 60
      - type: Percent
        value: 100
        periodSeconds: 60
      selectPolicy: Max
    
    # Conservative scale-down
    scaleDown:
      stabilizationWindowSeconds: 300  # 5 minutes
      policies:
      - type: Percent
        value: 25
        periodSeconds: 60
      - type: Pods
        value: 2
        periodSeconds: 60
      selectPolicy: Min
  
  # Metrics based on GitLab job queue
  metrics:
  - type: Pods
    pods:
      metric:
        name: gitlab_jobs_pending
      target:
        type: AverageValue
        averageValue: "1"

# .gitlab-ci.yml - Pipeline ตัวอย่างสำหรับใช้กับ Autoscaling Runner
stages:
  - build
  - test
  - security
  - deploy

variables:
  # Use Helm to deploy
  DEPLOYMENT_STRATEGY: helm
  # Enable debug logging
  CI_DEBUG_TRACE: "false"
  # Cache settings
  CACHE_KEY: "\${CI_COMMIT_REF_NAME}"
  DOCKER_TLS_CERTDIR: "/certs"

# Build stage
build:
  stage: build
  image: docker:24.0
  services:
    - docker:24.0-dind
  script:
    - docker build -t \${CI_REGISTRY_IMAGE}:\${CI_COMMIT_SHA} .
    - docker push \${CI_REGISTRY_IMAGE}:\${CI_COMMIT_SHA}
  artifacts:
    paths:
      - docker-image.tar
    expire_in: 1 hour

# Test stage
test:
  stage: test
  image: node:20-alpine
  script:
    - npm ci
    - npm test
  cache:
    key: "\${CI_COMMIT_REF_NAME}"
    paths:
      - node_modules/
      - .npm/

# Security scan
security-scan:
  stage: security
  image: aquasec/trivy:latest
  script:
    - trivy image \${CI_REGISTRY_IMAGE}:\${CI_COMMIT_SHA}
  allow_failure: true

# Deploy to Kubernetes
deploy:
  stage: deploy
  image: bitnami/kubectl:latest
  script:
    - kubectl set image deployment/my-app my-app=\${CI_REGISTRY_IMAGE}:\${CI_COMMIT_SHA}
    - kubectl rollout status deployment/my-app
  environment:
    name: production
    url: https://app.example.com
  only:
    - main

# GitLab Runner Deployment
apiVersion: apps/v1
kind: Deployment
metadata:
  name: gitlab-runner
  namespace: gitlab-runners
  labels:
    app: gitlab-runner
spec:
  replicas: 1
  selector:
    matchLabels:
      app: gitlab-runner
  template:
    metadata:
      labels:
        app: gitlab-runner
    spec:
      serviceAccountName: gitlab-runner
      securityContext:
        runAsNonRoot: true
        fsGroup: 65534
      
      containers:
      - name: gitlab-runner
        image: gitlab/gitlab-runner:latest
        securityContext:
          allowPrivilegeEscalation: false
          readOnlyRootFilesystem: true
        env:
        - name: CI_SERVER_URL
          value: "https://gitlab.example.com/"
        - name: REGISTRATION_TOKEN
          valueFrom:
            secretKeyRef:
              name: gitlab-runner-token
              key: registration-token
        - name: RUNNER_EXECUTOR
          value: "kubernetes"
        resources:
          requests:
            cpu: "100m"
            memory: "256Mi"
          limits:
            cpu: "500m"
            memory: "1Gi"
        volumeMounts:
        - name: config
          mountPath: /etc/gitlab-runner
        - name: cache
          mountPath: /cache
        - name: docker-config
          mountPath: /certs/client
      volumes:
      - name: config
        configMap:
          name: gitlab-runner-config
      - name: cache
        emptyDir: {}
      - name: docker-config
        emptyDir: {}

# ตัวอย่าง Jobs ที่พบบ่อยและวิธีแก้ไข

# ปัญหา: Runner ไม่ scale up
# แก้ไข: เพิ่ม time_window และ scale_up_delay
[runners.autoscale]
  time_window = "120s"  # เพิ่มเวลาตรวจสอบ jobs
  scale_up_delay = "30s"  # Delay ก่อน scale up

# ปัญหา: Jobs ค้างใน queue
# แก้ไข: ตรวจสอบ resources ของ runner pods
limits:
  memory: "2Gi"
  cpu: "1000m"
requests:
  memory: "512Mi"
  cpu: "500m"

# ปัญหา: Cache miss บ่อย
# แก้ไข: ใช้ S3 cache แทน default
[runners.cache]
  Type = "s3"
  Path = "gitlab-runner/cache"
  Shared = true
  [runners.cache.s3]
    ServerAddress = "s3.amazonaws.com"
    AccessKey = "YOUR_ACCESS_KEY"
    SecretKey = "YOUR_SECRET_KEY"
    BucketName = "gitlab-cache"