DNS Migration Checklist for Faster Websites: TTL Planning, Cutover, and Verification
DNSDomain MigrationTTLDNSSECHosting MigrationReliability

DNS Migration Checklist for Faster Websites: TTL Planning, Cutover, and Verification

CCaching.website Editorial Team
2026-11-15
8 min read

Use this DNS migration checklist to plan TTLs, verify records, handle DNSSEC, cut over safely, and maintain a practical rollback path.

A DNS migration does not need to cause downtime, broken email, or slower page loads. This checklist explains how to plan TTL changes, prepare authoritative nameservers, handle DNSSEC and CNAME flattening, verify the cutover across resolvers, and maintain a rollback path after moving a domain, host, or CDN.

Overview

Changing DNS providers or pointing a domain to a new web host is a controlled infrastructure change, not simply an edit to an A record. DNS connects visitors and services to your website, email platform, APIs, verification endpoints, monitoring tools, and sometimes CDN edge locations. A migration is successful when each required record remains available, the new destination answers correctly, and the old configuration can be restored if a problem appears.

The safest approach is to separate preparation from cutover. First, inventory the current zone and test the new hosting or CDN destination without changing public DNS. Next, reduce relevant TTL values before the change, confirm nameserver and DNSSEC details, and record a rollback configuration. After the change, verify authoritative answers and recursive resolver behavior from several networks. Finally, monitor the application, certificates, email, and origin traffic for at least one complete operational cycle before removing the old setup.

DNS TTL settings control how long resolvers may retain an answer. They do not force every resolver to refresh at the same moment, and they do not control the time required for a nameserver delegation change to become visible everywhere. This distinction matters: changing records inside a zone and changing the domain's authoritative nameservers are related but different migration events.

For broader preparation, pair this checklist with a website hosting migration performance checklist and a hosting performance benchmark template. DNS can direct traffic to a faster server, but it cannot compensate for an underperforming origin or an incorrectly configured cache.

What to track

1. Build a complete DNS inventory

Export or document the current zone before making changes. Include the record name, type, value, TTL, purpose, and owner. At minimum, review:

  • A and AAAA records: website and API destinations for IPv4 and IPv6.
  • CNAME records: CDN, hosted service, verification, and environment aliases.
  • MX, SPF, DKIM, and DMARC records: email delivery and authentication dependencies.
  • TXT records: domain verification, security policies, and service integrations.
  • CAA records: certificate issuance restrictions, if used.
  • NS and DNSSEC records: delegation, signing, and validation dependencies.

Record which systems depend on each entry. A web migration that preserves the homepage but drops an email or verification record is not complete. Also identify records managed outside the main DNS interface, such as registrar defaults, email-provider instructions, or automated infrastructure tooling.

2. Decide what is moving

There are three common scenarios:

  • Changing a destination: You keep the same authoritative DNS provider and update records such as A, AAAA, or CNAME.
  • Changing DNS providers: You recreate the entire zone at a new provider, then change the domain's authoritative nameservers at the registrar.
  • Changing a host or CDN: You may update DNS records, origin settings, certificates, cache rules, and health checks together.

Do not assume that a provider automatically copies every record. Compare the old and new zone line by line. Pay particular attention to the root domain, often represented as @, because a root CNAME may require provider-specific CNAME flattening or an equivalent alias feature.

3. Plan TTL changes

Before a planned cutover, lower the TTL on records that will change. The exact timing depends on the current TTL, resolver behavior, and your risk tolerance. A practical planning table is:

Migration itemBefore cutoverAfter verification
A, AAAA, or CNAME records being changedLower the TTL ahead of the maintenance window and confirm the lower value is being served.Restore a suitable operational TTL after the rollback window ends.
Records that will not changeUsually leave them unchanged unless the provider migration requires a full zone transition.Confirm they were copied accurately.
Authoritative nameserversPrepare the new zone and registrar change; do not treat record TTL reduction as a guarantee of instant delegation changes.Monitor authoritative answers and delegation status across resolvers.

Changing TTL after a value has already been cached does not necessarily shorten that existing cache period. Lowering TTL in advance gives resolvers an opportunity to receive the shorter value before the actual destination changes.

4. Prepare DNSSEC, CNAME flattening, and health checks

DNSSEC requires a specific transition plan. Identify whether the current zone is signed, which provider holds the signing keys, and whether the registrar has a DS record for the domain. A mismatch between delegation, signatures, and DS data can make validating resolvers reject the zone. Follow the receiving provider's documented DNSSEC migration sequence and define a rollback procedure before changing the delegation.

Confirm how the new provider handles the zone apex. CNAME flattening can return address records for a root domain while retaining an alias behind the scenes, but implementation and interface names vary. Test the actual authoritative response rather than assuming that an entered record behaves like a conventional CNAME.

If the provider offers health checks or automatic failover, define the monitored endpoint, expected status, timeout behavior, and fallback destination. A health check should test an endpoint that represents real service availability, not merely whether a server accepts a connection. For origin failover planning, see the CDN origin failover runbook.

5. Create a verification and rollback record

Save the old zone, registrar settings, DNSSEC details, certificate information, and the intended new values in a versioned document. Include the planned change time, responsible operator, validation commands, and rollback decision criteria. Keep the old hosting or origin available until traffic, application logs, email, and monitoring confirm that the new path is stable.

Cadence and checkpoints

Use these checkpoints for every migration, whether it involves a web host, CDN, domain registrar, or DNS provider.

  1. Several days before: Inventory records, identify dependencies, lower relevant TTLs, and benchmark the current website. Capture DNS lookup time, time to first byte, response status, certificate validity, and representative page performance.
  2. Before the change window: Build the new zone, validate all records, provision the certificate, test the new origin directly where possible, and confirm cache and host-header behavior. If caching changes too, consult the CDN migration checklist.
  3. Immediately before cutover: Confirm backups, access to the registrar and DNS provider, rollback authority, maintenance contacts, and monitoring alerts. Freeze unrelated DNS edits.
  4. At cutover: Apply the smallest required change. Note the exact time and query the authoritative nameservers directly to confirm the intended answer.
  5. During the first hour: Test the root domain, www host, important subdomains, HTTP-to-HTTPS redirects, IPv4 and IPv6, login paths, checkout flows, APIs, and email-related records. Check logs for requests reaching both old and new destinations.
  6. After the initial window: Query from multiple recursive resolvers and networks. Review uptime, origin errors, TLS alerts, cache behavior, mail delivery, and customer reports. Do not remove the old environment until the defined rollback window has passed.
  7. Monthly or quarterly: Recheck authoritative nameservers, DNSSEC status, expiring certificates, unused records, record ownership, and observed DNS performance. Update the inventory when infrastructure changes.

For repeatable operations, store the checklist with infrastructure code or the team's change-management records. A simple spreadsheet is sufficient if it clearly shows the prior value, new value, TTL, verification result, and change owner.

How to interpret changes

DNS propagation verification should compare several layers rather than relying on one online checker.

  • Authoritative response: Query each authoritative nameserver directly. If they disagree, the zone or delegation may be incomplete.
  • Recursive response: Query multiple public and local resolvers. Different answers during the transition can be expected while cached responses expire.
  • Application response: Resolve the hostname, connect over HTTPS, inspect the certificate, and request important pages. A correct DNS answer does not prove that the application is ready.
  • Regional behavior: Test from more than one network or location, especially if the domain uses CDN, geo-routing, or health-based answers.

If some resolvers show the old address while authoritative servers show the new one, cached data is a likely explanation; continue monitoring against the planned TTL and migration window. If authoritative servers disagree, investigate the zone, nameserver delegation, or provider synchronization before waiting. If DNS is correct but the page is slow, inspect origin latency, TLS negotiation, CDN routing, and cache headers. DNS performance is only one part of website speed optimization.

Unexpected results can also come from split-horizon DNS, local resolver overrides, stale application configuration, IPv6 pointing to a different server, or an old CDN cache. Check both A and AAAA answers and test with a clean resolver. If content is wrong or stale after the destination is correct, review cache keys and invalidation using the CDN cache keys guide.

Rollback when the new path causes material availability, security, transaction, or email problems and the cause cannot be corrected within the agreed window. Restore the documented prior value or delegation, verify authoritative answers, and continue monitoring cached responses. Do not make repeated alternating changes without recording each one; that can make diagnosis harder and extend the transition.

When to revisit

Revisit this DNS migration checklist before every host, CDN, registrar, or authoritative DNS provider change. It is also useful before certificate renewals, major domain changes, IPv6 enablement, email-provider moves, disaster-recovery exercises, and changes to health-based routing.

Schedule a monthly or quarterly review even when no migration is planned. Confirm that the nameservers still match the intended provider, DNSSEC remains valid, important records have an owner, TTLs reflect operational needs, and monitoring can detect resolution failures. Compare DNS lookup and website response measurements with the baseline captured before the last infrastructure change.

Refresh the procedure when a provider changes its terminology, DNSSEC workflow, CNAME flattening behavior, health-check options, or testing tools. Interfaces and features change, so keep screenshots and provider-specific instructions separate from this evergreen workflow. The durable record should remain the inventory, decision log, verification evidence, and rollback plan.

Before your next cutover, complete these final actions:

  • Export the current zone and document every dependency.
  • Lower only the TTLs that need to change, early enough to be useful.
  • Validate the new zone, DNSSEC plan, apex behavior, certificates, and health checks.
  • Record the old configuration and define clear rollback criteria.
  • Test authoritative answers, recursive responses, application behavior, email, and IPv4/IPv6 paths.
  • Keep monitoring after the cutover and restore normal TTLs only after stability is demonstrated.

Related Topics

#DNS#Domain Migration#TTL#DNSSEC#Hosting Migration#Reliability
C

Caching.website Editorial Team

Technical SEO 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.