Website Migration Runbook: A No-Downtime Workflow for Moving Hosts or Platforms
website-migrationhosting-changedns-cutoverstagingrollback

Website Migration Runbook: A No-Downtime Workflow for Moving Hosts or Platforms

CCloudSite Studio Editorial Team
2026-11-18
7 min read

Use this practical website migration runbook to move hosts or platforms with tested backups, staged verification, DNS controls, and rollback criteria.

Moving a website to a new host or platform does not have to mean lost traffic, broken email, or an extended outage. This website migration runbook gives you a repeatable sequence for inventory, backups, staging, testing, DNS cutover, rollback, and first-week monitoring so you can move with controlled risk.

Overview

A no-downtime website migration is best treated as a controlled release rather than a single hosting change. The existing website remains available while you prepare and test a complete copy on the new environment. Only after the copy passes functional, technical, and business checks do you direct visitors to it.

The workflow applies when you are moving to a new web host, changing server architecture, migrating to cloud hosting, replacing a content management system, or switching between a hosted website builder and a self-managed platform. The exact control-panel labels will vary, but the checkpoints remain broadly useful.

Before beginning, assign an owner for the migration and record the planned maintenance window, even if you expect no visible downtime. Create a shared log with these fields:

  • Date and time of each action
  • Person responsible
  • What was changed
  • Evidence or test result
  • Decision: proceed, pause, or roll back

Also record the current host, new host, domain registrar, DNS provider, email provider, database type, CMS and version, SSL arrangement, analytics tools, forms, payment integrations, scheduled jobs, and third-party services. This inventory prevents a common migration failure: moving the pages while overlooking a dependent service.

What to track

1. Backups and recovery points

Take a complete backup before changing anything. Include website files, databases, uploaded media, configuration files, certificates or certificate settings where applicable, and any export available from a website builder. Keep at least one copy outside the current hosting account and verify that it can be opened or restored.

Do not treat a backup as proven until you have checked it. Confirm that the database export contains recent records, the media directory has expected files, and the archive is not empty or corrupt. If the website handles orders, bookings, comments, or contact submissions, record the latest transaction or submission so you can compare it after the move.

2. URLs, redirects, and search signals

Export or crawl the current URL list. Include important pages, images, downloadable files, blog posts, canonical URLs, existing redirects, and XML sitemap locations. Note the status code and final destination for old URLs that must continue to work.

Preserve URL structures where possible. If URLs change, map every important old address to its most relevant new address with a permanent server-side redirect. Avoid sending all old URLs to the home page, because that can create a poor visitor experience and obscure which content moved successfully.

3. Email and DNS dependencies

List every DNS record before editing the zone: A and AAAA records, CNAME records, MX records, TXT records, verification records, and any subdomains. Email records are especially important. A website migration can leave the site working while business email stops receiving messages if MX, SPF, DKIM, or DMARC-related records are overwritten.

Use the existing DNS zone as the reference, not the new host's default template. For more detail on mail continuity, review the business email setup and deliverability guide.

4. Performance, security, and functionality

Record a small baseline before migration: representative page load behavior, uptime observations, key response codes, current SSL status, and important form or checkout outcomes. Test the new environment for the same items. Confirm that updates, backups, access controls, firewall settings, and SSL renewal are configured rather than assuming the new host supplied them automatically.

Cadence and checkpoints

Phase one: prepare the inventory

Complete the inventory, lower DNS record time-to-live where appropriate, and schedule the cutover during a period when your team can monitor the site. A lower TTL may help resolvers request updated records sooner, but it does not guarantee an instant change. Resolver caches and provider behavior vary, so plan for a transition period.

Phase two: build and stage

Create the destination environment without changing public DNS. Copy files and databases using the method supported by your platform: a host migration tool, CMS export and import, secure file transfer, database dump and restore, or a builder-specific transfer process. Keep the source site available while this copy is prepared.

Give the staging copy a private address or access restriction. Do not allow a test copy to be indexed as a duplicate of the live site. If the site uses environment-specific settings, update database credentials, file paths, API keys, cron jobs, caching, and payment or email integrations carefully.

Phase three: verify before cutover

Test the staging site from its temporary address or through a local hosts-file override. Check:

  • Home page, navigation, search, categories, and pagination
  • Forms, notifications, logins, account areas, bookings, and checkout flows
  • Images, downloads, embedded media, and responsive layouts
  • Database content, recent records, comments, and user accounts
  • Canonical tags, robots directives, sitemap, metadata, and redirects
  • Analytics, tag management, consent settings, and conversion tracking
  • SSL, mixed-content warnings, error pages, and security headers where configured

Compare the staging copy with the source using a URL checklist and sample records. Fix failures before cutover; DNS changes should not be used as a testing method.

Phase four: cut over

Take a final backup and, if the site accepts changing data, place it in a short maintenance or read-only mode if the platform supports that option. Perform a final database synchronization so new submissions are not lost. Update only the records needed to point the domain to the new environment, leaving mail and unrelated subdomains unchanged.

After saving DNS changes, check the domain from several networks or DNS lookup tools. Keep the old host active until the new site is confirmed and the expected propagation period has passed. Do not cancel the old account immediately.

Phase five: monitor and close

Run the same functional tests on the live domain. Check the homepage, a deep content page, a form, a recently changed record, a redirect, the SSL certificate, analytics collection, and email delivery. Review server logs, application logs, search-console-style coverage reports if available, uptime alerts, and error monitoring.

How to interpret changes

Separate DNS symptoms from application problems. If some networks reach the old server and others reach the new one, propagation or caching may still be in progress. If all visitors receive a 500 error, blank page, database connection error, or redirect loop, investigate the destination environment rather than waiting for DNS.

A small number of missing images may indicate an incomplete file transfer or case-sensitive path differences. Missing form submissions may point to changed mail credentials, blocked outbound mail, incorrect environment variables, or a form endpoint that still references the old host. A sudden drop in tracked conversions can be caused by a missing analytics tag or consent configuration even when pages appear normal.

Use explicit rollback criteria. Consider reverting DNS or restoring the previous environment when the new site has widespread errors, critical transactions fail, data is missing, email is disrupted, SSL cannot be fixed promptly, or the team cannot identify the cause within the agreed window. Preserve logs and the new environment before rolling back so investigation can continue without destroying evidence.

Once stable, compare traffic, indexed URLs, crawl errors, response codes, uptime, form completions, and business transactions with the pre-migration baseline. A migration is not complete simply because the homepage loads.

When to revisit

Use this runbook before every host, CMS, domain, theme, or infrastructure move. Revisit it monthly during the first few months after a major migration, then at least quarterly as part of routine website maintenance. Review the monitoring log for recurring 4xx or 5xx errors, redirect chains, slow pages, failed backups, certificate warnings, missing analytics events, and unexpected DNS changes.

Update the runbook whenever your hosting control panel, CMS version, deployment process, DNS provider, email provider, or security tools change. Platform-specific instructions can become outdated even when the migration principles remain sound. Refresh the backup and restore procedure after testing it, and repeat a restore test on a schedule appropriate to the site's risk.

For the next migration, duplicate your log, replace the inventory with current values, and keep a short decision record for every checkpoint. Pair this process with the website backup and restore guide, the DNS troubleshooting guide, and the uptime monitoring guide. A documented baseline, tested recovery point, staged copy, careful DNS cutover, and first-week review provide a practical path to moving a website without sacrificing its data, traffic, or email.

Related Topics

#website-migration#hosting-change#dns-cutover#staging#rollback
C

CloudSite Studio Editorial Team

Web Hosting and Site Building Editors

Senior editor and content strategist. Writing about technology, design, and the future of digital media. Follow along for deep dives into the industry's moving parts.