database.yaml

name: Database

on:
  workflow_dispatch:
  schedule:
    - cron: "0 4 * * *" # 04:00 UTC

env:
  SERVICE_NAME: apply-for-qualified-teacher-status
  SERVICE_SHORT: afqts
  TF_VARS_PATH: terraform/application/config

jobs:
  backup-production:
    name: Backup database
    runs-on: ubuntu-latest
    environment:
      name: ${{ inputs.environment || 'production' }}
    env:
      DEPLOY_ENV: ${{ inputs.environment || 'production'  }}
      BACKUP_FILE: ${{ inputs.backup-file || 'schedule'  }}

    services:
      postgres:
        image: postgres:14
        env:
          POSTGRES_USER: postgres
          POSTGRES_PASSWORD: postgres
          POSTGRES_DB: postgres
        ports:
          - 5432:5432
        options: --health-cmd pg_isready --health-interval 10s --health-timeout 5s --health-retries 5

    steps:
      - name: Checkout code
        uses: actions/checkout@v4

      - uses: azure/login@v2
        with:
          creds: ${{ secrets.AZURE_CREDENTIALS }}

      - name: Set environment variables
        run: |
          source global_config/${DEPLOY_ENV}.sh
          tf_vars_file=${TF_VARS_PATH}/${DEPLOY_ENV}/variables.tfvars.json
          echo "CLUSTER=$(jq -r '.cluster' ${tf_vars_file})" >> $GITHUB_ENV
          echo "RESOURCE_GROUP_NAME=${AZURE_RESOURCE_PREFIX}-${SERVICE_SHORT}-${CONFIG_SHORT}-rg" >> $GITHUB_ENV
          echo "STORAGE_ACCOUNT_NAME=${AZURE_RESOURCE_PREFIX}${SERVICE_SHORT}dbbkp${CONFIG_SHORT}sa" >> $GITHUB_ENV
          TODAY=$(date +"%F")
          echo "DB_SERVER=${AZURE_RESOURCE_PREFIX}-${SERVICE_SHORT}-${CONFIG_SHORT}-pg" >> $GITHUB_ENV
          if [ "${BACKUP_FILE}" == "schedule" ]; then
            BACKUP_FILE=${SERVICE_SHORT}_${CONFIG_SHORT}_${TODAY}
          elif [ "${BACKUP_FILE}" == "default" ]; then
            BACKUP_FILE=${SERVICE_SHORT}_${CONFIG_SHORT}_adhoc_${TODAY}
          else
            BACKUP_FILE=${BACKUP_FILE}
          fi
          echo "BACKUP_FILE=${BACKUP_FILE}" >> $GITHUB_ENV
          echo "KEYVAULT_NAME=${AZURE_RESOURCE_PREFIX}-${SERVICE_SHORT}-${CONFIG_SHORT}-inf-kv" >> $GITHUB_ENV

      - name: Fetch secrets from key vault
        uses: azure/CLI@v2
        id: key-vault-secrets
        with:
          inlineScript: |
            SLACK_WEBHOOK=$(az keyvault secret show --name "SLACK-WEBHOOK" --vault-name ${KEYVAULT_NAME} --query "value" -o tsv)
            echo "::add-mask::$SLACK_WEBHOOK"
            echo "SLACK_WEBHOOK=$SLACK_WEBHOOK" >> $GITHUB_OUTPUT

      - name: Backup ${{ env.DEPLOY_ENV }} postgres
        uses: DFE-Digital/github-actions/backup-postgres@master
        with:
          storage-account: ${{ env.STORAGE_ACCOUNT_NAME }}
          resource-group: ${{ env.RESOURCE_GROUP_NAME }}
          app-name: ${{ env.SERVICE_NAME }}-${{ env.DEPLOY_ENV }}
          cluster: ${{ env.CLUSTER }}
          azure-credentials: ${{ secrets.AZURE_CREDENTIALS }}
          backup-file: ${{ env.BACKUP_FILE }}.sql
          db-server-name: ${{ inputs.db-server }}
          slack-webhook: ${{ steps.key-vault-secrets.outputs.SLACK_WEBHOOK }}

      - uses: ./.github/actions/set-kubernetes-credentials
        with:
          environment: production
          azure-credentials: ${{ secrets.AZURE_CREDENTIALS }}

      - name: Refresh Terraform
        run: make ci production terraform-refresh
        env:
          TF_VAR_azure_sp_credentials_json: ${{ secrets.AZURE_CREDENTIALS }}
          DOCKER_IMAGE: "ghcr.io/dfe-digital/apply-for-qualified-teacher-status:no-tag"

      - name: Get Azure backup storage details
        id: azure-backup-storage
        working-directory: terraform/application
        run: |
          echo "account-name=$(terraform output -raw postgres_azure_backup_storage_account_name)" >> $GITHUB_OUTPUT
          echo "container-name=$(terraform output -raw postgres_azure_backup_storage_container_name)" >> $GITHUB_OUTPUT

      - name: Install postgres client
        uses: DFE-Digital/github-actions/install-postgres-client@master
        with:
          version: 14

      - name: Sanitise dump
        run: |
          gzip -d --to-stdout backup.sql.gz | psql -d postgres
          psql -d postgres -f db/scripts/sanitise.sql
          pg_dump -E utf8 --compress=1 --clean --if-exists --no-owner --verbose --no-password -f backup-sanitised.sql.gz
        env:
          PGUSER: postgres
          PGPASSWORD: postgres
          PGHOST: localhost
          PGPORT: 5432

      - name: Upload sanitised backup
        uses: actions/upload-artifact@v4
        with:
          name: backup-sanitised
          path: backup-sanitised.sql.gz
          retention-days: 3

  restore-preproduction:
    name: Restore preproduction
    needs: [backup-production]
    runs-on: ubuntu-latest

    environment: preproduction

    steps:
      - name: Checkout code
        uses: actions/checkout@v4

      - name: Download sanitised backup
        uses: actions/download-artifact@v4
        with:
          name: backup-sanitised

      - uses: ./.github/actions/set-kubernetes-credentials
        with:
          environment: preproduction
          azure-credentials: ${{ secrets.AZURE_CREDENTIALS }}

      - name: Install kubectl
        uses: DFE-Digital/github-actions/set-kubectl@master

      - name: Install konduit
        run: make install-konduit

      - name: Restore sanitised backup
        shell: bash
        run: bin/konduit.sh -i backup-sanitised.sql.gz -c apply-for-qts-preproduction-web -- psql