Added job to backfill ghost_uuid in sites table (#1432)

Added job to backfill `ghost_uuid` in `sites` table

ref https://linear.app/ghost/issue/BER-2964

Added job that backfills the `ghost_uuid` field recently added to the `sites`
table

Author: mike182uk

biome.json

@@ -23,7 +23,7 @@
             "groups": [
               ["vitest", "@cucumber/cucumber"],
               ":BLANK_LINE:",
-              ":NODE:",
+              [":NODE:", ":BUN:"],
               ":BLANK_LINE:",
               ":PACKAGE:",
               ":BLANK_LINE:",

jobs/__template__/index.test.ts

@@ -1,5 +1,3 @@
-import mysql from 'mysql2/promise';
-
 import {
     afterAll,
     beforeAll,
@@ -8,6 +6,9 @@ import {
     expect,
     test,
 } from 'bun:test';
+
+import mysql from 'mysql2/promise';
+
 import { doSomething } from './index';
 
 describe('__JOB_NAME__', () => {

jobs/backfill-ghost-uuid/Dockerfile

@@ -0,0 +1,8 @@
+FROM oven/bun:1.3.0-alpine@sha256:37e6b1cbe053939bccf6ae4507977ed957eaa6e7f275670b72ad6348e0d2c11f
+
+WORKDIR /opt/job
+
+COPY index.ts .
+
+ENTRYPOINT ["bun", "run", "index.ts"]
+CMD []

jobs/backfill-ghost-uuid/README.md

@@ -0,0 +1,67 @@
+# backfill-ghost-uuid
+
+Backfills the `ghost_uuid` field in the `sites` table by fetching the UUID from 
+each site's `/ghost/api/admin/site/` endpoint
+
+For each `site` in the database without a `ghost_uuid`, this job:
+
+1. Makes a request to `https://<site_host>/ghost/api/admin/site/`
+2. Extracts the `site_uuid` from the response
+3. Updates the site record in the database
+
+Sites are processed sequentially with a 500ms delay between requests to avoid 
+tripping rate limits. If a request fails, the job logs a warning and continues 
+to the next site
+
+## Prerequisites
+
+- Globally available `bun`
+- Globally available `docker`
+
+## Running locally
+
+```bash
+DB_HOST=... \
+DB_PORT=... \
+DB_USER=... \
+DB_PASSWORD=... \
+DB_NAME=... \
+bun run index.ts
+```
+
+## Running tests
+
+```bash
+./run-tests.sh
+```
+
+## Production deployment
+
+Build and push the Docker image to GCP:
+
+```bash
+./gcloud-push.sh
+```
+
+## Production execution
+
+Use the [GCP console](https://console.cloud.google.com/run/jobs/create) to setup and execute the job
+
+Ensure the relevant environment variables are set on the job:
+
+```text
+DB_HOST
+DB_PORT
+DB_USER
+DB_PASSWORD
+DB_NAME
+```
+
+If using a socket connection, set `DB_SOCKET_PATH` instead of `DB_HOST` and `DB_PORT`
+
+## Notes
+
+- The job is idempotent and can be run multiple times safely
+- Sites are processed sequentially to avoid tripping rate limits
+- Failed requests are logged but don't stop the job from processing remaining sites
+- The job is designed to be run periodically to catch any sites that failed on previous runs

jobs/backfill-ghost-uuid/docker-compose.yml

@@ -0,0 +1,10 @@
+services:
+  mysql-test:
+    image: mysql:8.4.6@sha256:5367102acfefeaa47eb0eb57c8d4f8b96c8c14004859131eac9bbfaa62f81e34
+    environment:
+      MYSQL_ROOT_PASSWORD: root
+      MYSQL_DATABASE: backfill-ghost-uuid
+    ports:
+      - "3308:3306"
+    tmpfs:
+      - /var/lib/mysql

jobs/backfill-ghost-uuid/gcloud-push.sh

@@ -0,0 +1,39 @@
+#!/usr/bin/env bash
+
+set -euo pipefail
+
+IMAGE_NAME="backfill-ghost-uuid-job"
+
+PROJECT_ID=$(gcloud config get-value project 2>/dev/null)
+if [ -z "$PROJECT_ID" ]; then
+  echo "Error: Could not determine GCP project ID"
+  exit 1
+fi
+
+echo ""
+echo "Building and pushing to GCP..."
+echo ""
+echo "Project: $PROJECT_ID"
+echo "Image:   $IMAGE_NAME"
+echo ""
+
+echo "Building Docker image..."
+echo ""
+docker build --platform linux/amd64 -t "$IMAGE_NAME" .
+echo ""
+
+IMAGE_URL="gcr.io/$PROJECT_ID/$IMAGE_NAME:latest"
+echo "Tagging image: $IMAGE_URL"
+docker tag "$IMAGE_NAME" "$IMAGE_URL"
+echo ""
+
+echo "Pushing image to Google Container Registry..."
+echo ""
+docker push "$IMAGE_URL"
+echo ""
+
+echo ""
+echo "Push complete!"
+echo ""
+echo "Built and pushed image to '$IMAGE_URL'"
+echo ""