Skip to content

README.md

Latest commit

 

History

History
151 lines (105 loc) · 3.52 KB

README.md

File metadata and controls

151 lines (105 loc) · 3.52 KB

Supabase to Cloudflare Migration Guide

This document records the one-time migration from Supabase to Cloudflare for NameUtils.

Migration Summary

  • Frontend stays on Next.js
  • Authentication moved from Supabase Auth to Auth.js + Google
  • Primary data moved from Supabase Postgres to Cloudflare D1
  • Cache moved to Cloudflare KV
  • Migration inputs, reports, and backups moved to Cloudflare R2

Data Mapping

Old platform

  • Supabase Auth users
  • domains
  • domain_tags
  • whois_cache

New platform

  • users in D1
  • domains in D1
  • domain_tags in D1
  • WHOIS cache in KV

whois_cache is no longer treated as primary data and is not imported into D1.

Required Cloudflare Resources

  • 1 D1 database
  • 1 KV namespace
  • 1 R2 bucket
  • 1 Worker deployment for the Next.js app

Auth Changes

  • Google OAuth callback:
https://nameutils.com/api/auth/callback/google
  • Optional additional callbacks:
https://www.nameutils.com/api/auth/callback/google
http://localhost:3000/api/auth/callback/google

Migration Inputs

Depending on the export you have, use one of these two paths:

Full Supabase export

Use:

  • old user ID to email mapping
  • domains export
  • domain_tags export

Generate import artifacts:

node scripts/prepare-supabase-import.mjs \
  --user-map ./exports/user-id-map.json \
  --domains ./exports/domains.json \
  --domain-tags ./exports/domain-tags.json \
  --out-dir ./migration-artifacts

App-level domain export

Use this when you only have exported domain lists from the old app.

Generate import artifacts:

node scripts/prepare-domain-export-import.mjs \
  --email you@example.com \
  --export ./domains_export.json \
  --export "./domains_export (1).json" \
  --export "./domains_export (2).json" \
  --export "./domains_export (3).json" \
  --out-dir ./migration-artifacts/domain-export-import

This flow merges duplicate rows, keeps the most complete record for each domain, and generates a D1-ready SQL import file.

Generated Artifacts

The migration scripts produce:

  • users.json
  • domains.json
  • domain-tags.json
  • migration-report.json
  • d1-import.sql

Some flows also produce:

  • merge-details.json

Import into D1

  1. Apply the schema:
pnpm wrangler d1 migrations apply nameutils --remote
  1. Import the generated SQL:
pnpm wrangler d1 execute nameutils --remote --file=./migration-artifacts/d1-import.sql

If you used the app-level export flow, replace the path with the generated directory you chose.

Backup to R2

After review, you can upload the migration files to R2 for record keeping:

pnpm wrangler r2 object put nameutils-migration-assets/migration-report.json --remote --file=./migration-artifacts/migration-report.json
pnpm wrangler r2 object put nameutils-migration-assets/d1-import.sql --remote --file=./migration-artifacts/d1-import.sql

Post-Migration Checks

  • Google login succeeds
  • Protected pages require login
  • The expected domains appear under the correct account
  • Create, edit, delete, favorite, and import all work
  • Domain search works
  • WHOIS lookup works
  • Repeated search and WHOIS requests hit KV cache
  • No Supabase URL or public Supabase key is exposed in the browser

Rollout Order

  1. Finish migration testing in preview or staging
  2. Freeze writes on the old Supabase-backed app
  3. Run the final export
  4. Import the final dataset into D1
  5. Update secrets and production deployment
  6. Verify production behavior
  7. Keep the old system read-only for a short fallback period