hosting

The DNS Move That Prevents Website and Email Surprises

Changing hosting providers looks simple until a website loads for some visitors, email keeps arriving for others, and nobody can explain which DNS zone is actually authoritative. The risky part is rarely the new…

Featured image for The DNS Move That Prevents Website and Email Surprises

Changing hosting providers looks simple until a website loads for some visitors, email keeps arriving for others, and nobody can explain which DNS zone is actually authoritative. The risky part is rarely the new server itself. It is the gap between what you copied, what resolvers cached, and what your domain’s nameservers are currently publishing.

A controlled DNS move is a staged migration. You inventory the old zone, build the new one before switching traffic, shorten the right TTLs, test both web and mail records, and keep a rollback path until the new setup has proved stable. The sequence below is designed for a small business site, WordPress installation, SaaS project, or agency-managed domain.

1. Map the current zone before changing anything

Start by exporting or copying every record from the current DNS provider. Do not limit the list to the root website record. At minimum, look for:

  • A and AAAA records: the IPv4 and IPv6 destinations for the root domain.
  • CNAME records: common for www, app subdomains, verification, and hosted services.
  • MX records: the mail servers that receive messages for the domain.
  • TXT records: SPF, DKIM, DMARC, ownership checks, and service configuration.
  • CAA, SRV, and delegated subdomain records: less common, but important when present.

Also record each item’s name, value, priority where applicable, proxy or DNS-only setting, and TTL. A DNS export is useful, but read it rather than trusting an import button blindly: providers use different defaults, field names, and behaviours for blank hostnames and trailing dots.

2. Reduce TTLs before the migration window

TTL, or time to live, tells recursive resolvers how long they may reuse a cached DNS answer. AWS explains that shorter TTLs make changes visible sooner but can increase query volume, while longer TTLs reduce repeat lookups and are often more stable.

For a planned move, lower the TTL on the records you expect to change before you change them. The exact value depends on the provider and the risk you can tolerate; a short migration window may use a few minutes, while a cautious team may choose a longer value. The important limitation is timing: lowering TTL today does not erase answers that were cached under yesterday’s longer TTL.

Nameserver delegation has its own caching behaviour. If you are moving DNS service, not just the web server, check the provider’s guidance for the NS record and registrar delegation. Amazon Route 53’s migration procedure notes that nameserver information may otherwise remain cached for up to two days.

3. Build and verify the new zone before cutover

Create the new zone while the old provider is still serving production traffic. Reproduce the records first, then compare the two zones line by line. This is where many migrations fail: the new provider contains the website’s A record but misses an MX record, a DKIM selector, or a subdomain used by a payment, help-desk, analytics, or deployment service.

Use a deliberately boring test domain or a temporary hostname where possible. Check that:

  • the root domain and www resolve to the intended destination;
  • IPv6 does not send visitors to an old server;
  • the certificate and HTTPS redirect work on every public hostname;
  • MX, SPF, DKIM, and DMARC records still describe the intended mail flow;
  • verification records for third-party services are present and unchanged.

Remember that a DNS provider can host a correct zone without being authoritative for the domain yet. The registrar’s nameserver delegation determines which provider the public DNS system consults.

4. Separate the website move from the DNS move when possible

If you are changing both hosting and DNS providers on the same day, you create two moving parts. A lower-risk sequence is to prepare the new website first, test it through a temporary hostname or provider preview URL, and confirm that the application, uploads, database connection, redirects, and scheduled tasks work before changing DNS.

Then the DNS change only needs to answer one question: should this hostname point to the old host or the new host? If the new server is not ready, changing DNS earlier simply makes the failure public.

5. Choose the smallest cutover that proves the plan

For a web-only move, you may only need to change an A, AAAA, or CNAME record. Keep email records untouched if mail is staying with the existing provider. For a full DNS-service migration, create the complete zone first and then change the registrar’s nameservers to the new provider’s assigned nameservers.

Make the change during a monitored maintenance window in India Standard Time, with the person who controls the registrar account available. Save the old nameservers and the old zone export before clicking the final confirmation. A screenshot is not a substitute for a machine-readable copy, but it can help when reconciling an emergency rollback.

6. Monitor from more than one network

After cutover, test from multiple recursive resolvers and networks instead of relying only on the office connection. Check the site, login flow, forms, checkout or booking path, APIs, and email in both directions. Watch application logs for traffic still reaching the old host, but do not assume that every old request means the migration failed: cached DNS answers expire at different times.

Keep the old hosting account and old DNS zone available until the agreed observation period has passed. If the new setup is wrong, rollback is easier when the previous records and server are still intact. A rollback does not instantly reach every resolver either, so document the decision and communicate clearly.

A concrete example

Imagine example.com moving a WordPress site to a new host while keeping email at the current mail provider. The inventory contains an A record for @, a CNAME for www, two MX records, SPF and DKIM TXT records, a DMARC record, and a verification CNAME for a marketing platform.

The team prepares the new WordPress site, copies every record into the new zone, lowers the web record TTL ahead of the window, and tests the new host directly. At cutover, only the web destination changes. MX and mail-related TXT records remain identical. Afterward, the team checks both website and email delivery, monitors logs, and keeps the old host until traffic and support reports are normal.

Trade-offs worth remembering

Short TTLs improve responsiveness during planned changes, but they can increase DNS queries and do not control every cache in the path. A full nameserver move affects more records than a web-host move. DNSSEC adds another dependency: signing keys and delegation must be coordinated with the new provider rather than copied casually. Provider-specific features such as proxying, traffic steering, health checks, and alias records may not have exact equivalents elsewhere.

The best migration is therefore not the one with the most automation. It is the one where the team knows what is authoritative, what is cached, what is being changed, and how to return to the previous working state.

Final checklist: export the old zone; inventory web, mail, and verification records; lower relevant TTLs early; build the new zone; test the new host; confirm nameserver delegation; make the smallest cutover; monitor from multiple networks; and keep a rollback copy until the change is stable.

This article is a general information summary based on the cited public sources. It is not legal, cybersecurity, or compliance advice.

Written By

Shaik Sai Meera Editorial Team

shares practical publishing insights, comparisons, and WordPress-focused growth guidance for modern web teams.

136 articles published