zebra/.github/workflows/chore-delete-gcp-resources.yml

# This workflow is designed to delete old Google Cloud Platform (GCP) resources to save on costs.
#
# 1. Deletes specific instances in GCP older than a defined number of days.
# 2. Deletes instance templates older than a set number of days.
# 3. Deletes older disks not currently in use, with certain ones prefixed by commit hashes or "zebrad-".
# 4. Deletes cache images from GCP, retaining a specified number of the latest images for certain types like zebrad checkpoint cache, zebrad tip cache, and lightwalletd + zebrad tip cache.
# 5. Deletes unused artifacts from Google Artifact Registry older than a defined number of hours while retaining the latest few.
#
# It uses the gcloud CLI for most of its operations and also leverages specific GitHub Actions like the gcr-cleaner for deleting old images from the Google Artifact Registry.
# The workflow is scheduled to run daily at 0700 UTC.
name: Delete GCP resources

on:
  # Run daily, when most devs aren't working
  # 0700 UTC is after AEST working hours but before ET working hours
  schedule:
    - cron: "0 7 * * *"
  workflow_dispatch:

env:
  # Delete all resources created before $DELETE_INSTANCE_DAYS days ago.
  # We keep this short to reduce CPU, RAM, and storage costs.
  DELETE_INSTANCE_DAYS: 3
  # Delete all other resources created before $DELETE_AGE_DAYS days ago.
  # We keep this short to reduce storage costs.
  DELETE_AGE_DAYS: 2
  # But keep the latest $KEEP_LATEST_IMAGE_COUNT images of each type.
  # We keep this small to reduce storage costs.
  KEEP_LATEST_IMAGE_COUNT: 2
  # Delete all artifacts in registry created before $DELETE_IMAGE_HOURS hours ago.
  # We keep this long enough for PRs that are still on the same commit can re-run with the same image.
  DELETE_IMAGE_HOURS: 504h # 21 days

jobs:
  delete-resources:
    name: Delete old GCP resources
    runs-on: ubuntu-latest
    permissions:
      contents: 'read'
      id-token: 'write'
    steps:
      - uses: actions/checkout@v4.1.1
        with:
          persist-credentials: false

      # Setup gcloud CLI
      - name: Authenticate to Google Cloud
        id: auth
        uses: google-github-actions/auth@v2.1.2
        with:
          retries: '3'
          workload_identity_provider: '${{ vars.GCP_WIF }}'
          service_account: '${{ vars.GCP_DEPLOYMENTS_SA }}'

      - name: Set up Cloud SDK
        uses: google-github-actions/setup-gcloud@v2.1.0

      # Deletes all mainnet and testnet instances older than $DELETE_INSTANCE_DAYS days.
      #
      # We only delete instances that end in 7 or more hex characters,
      # to avoid deleting managed instance groups and manually created instances.
      #
      # ${INSTANCE_AND_ZONE} expands to:
      # <instance-name> --zone=<zone-name>
      # so it can't be shell-quoted.
      - name: Delete old instances
        run: |
          ./.github/workflows/scripts/gcp-delete-old-instances.sh
      # Deletes all the instance templates older than $DELETE_AGE_DAYS days.
      - name: Delete old instance templates
        run: |
          ./.github/workflows/scripts/gcp-delete-old-templates.sh

      # Deletes all mainnet and testnet disks older than $DELETE_AGE_DAYS days.
      #
      # Disks that are attached to an instance template can't be deleted, so it is safe to try to delete all disks here.
      #
      # ${DISK_AND_LOCATION} expands to:
      # <disk-name> --[zone|region]=<location-name>
      # so it can't be shell-quoted.
      - name: Delete old disks
        run: |
          ./.github/workflows/scripts/gcp-delete-old-disks.sh

      # Deletes mainnet and testnet cache images older than $DELETE_AGE_DAYS days.
      #
      # Keeps all images younger than $DELETE_AGE_DAYS.
      # Also keeps $KEEP_LATEST_IMAGE_COUNT older images of each type, for each network:
      # - zebrad checkpoint cache
      # - zebrad tip cache
      # - lightwalletd + zebrad tip cache
      #
      # TODO:
      # - refactor out repeated shell script code
      - name: Delete old cache images
        run: |
          ./.github/workflows/scripts/gcp-delete-old-cache-images.sh

  # We're using a generic approach here, which allows multiple registries to be included,
  # even those not related to GCP. Enough reason to create a separate job.
  #
  # The same artifacts are used for both mainnet and testnet.
  clean-registries:
    name: Delete unused artifacts in registry
    runs-on: ubuntu-latest
    permissions:
      contents: 'read'
      id-token: 'write'
    steps:
      - uses: actions/checkout@v4.1.1
        with:
          persist-credentials: false

      # Setup gcloud CLI
      - name: Authenticate to Google Cloud
        id: auth
        uses: google-github-actions/auth@v2.1.2
        with:
          retries: '3'
          workload_identity_provider: '${{ vars.GCP_WIF }}'
          service_account: '${{ vars.GCP_DEPLOYMENTS_SA }}'
          token_format: 'access_token'

      - name: Login to Google Artifact Registry
        uses: docker/login-action@v3.0.0
        with:
          registry: us-docker.pkg.dev
          username: oauth2accesstoken
          password: ${{ steps.auth.outputs.access_token }}

      # Deletes all images older than $DELETE_IMAGE_HOURS days.
      - uses: 'docker://us-docker.pkg.dev/gcr-cleaner/gcr-cleaner/gcr-cleaner-cli'
        continue-on-error: true # TODO: remove after fixig https://github.com/ZcashFoundation/zebra/issues/5933
        # Refer to the official documentation to understand available arguments:
        # https://github.com/GoogleCloudPlatform/gcr-cleaner
        with:
          args: >-
            -repo=us-docker.pkg.dev/${{ vars.GCP_PROJECT }}/zebra/zebrad-test
            -grace=${{ env.DELETE_IMAGE_HOURS }}
            -keep=${{ env.KEEP_LATEST_IMAGE_COUNT }}