When a website, subdomain, verification process, or mailbox stops working after a DNS change, the fastest solution is a structured check rather than repeated edits. This record-by-record DNS troubleshooting guide explains how to isolate propagation, incorrect values, nameserver mismatches, provider conflicts, and common configuration mistakes using a reusable lookup sequence and practical worksheets.
Overview
DNS translates human-readable hostnames into the destinations used by browsers, mail servers, and third-party services. A record can be correct in one control panel and still appear broken if the domain delegates DNS to a different provider, a resolver has cached an older answer, or the value does not match the service’s requirements.
Before changing anything, record the intended configuration. Write down the hostname, record type, value, TTL, the provider that should answer DNS queries, and the service that supplied the value. This simple DNS change worksheet prevents troubleshooting from turning into a series of undocumented edits.
- Hostname: for example, the root domain,
www,app, ormail. - Type: A, AAAA, CNAME, TXT, MX, or another required record.
- Value: the IP address, target hostname, verification string, or mail server.
- Expected provider: the DNS host whose nameservers are delegated at the registrar.
- Change time and TTL: useful when comparing authoritative and cached answers.
Start by checking delegation, then query the authoritative nameserver, and finally compare results from more than one public resolver. Tools such as dig, nslookup, and web-based DNS lookup tools can show whether the problem exists at the source or only in a cached response. If you are moving a site between a VPS, cloud host, or shared host, keep the old service available until the new DNS answers and application tests are complete. For broader migration planning, use the website migration checklist.
Checklist by scenario
1. A or AAAA record: the website does not load
- Confirm that the A record points to the current IPv4 address of the web server.
- Check for an AAAA record. Some clients prefer IPv6, so an outdated AAAA value can cause failures even when the A record is correct.
- Verify that the record is attached to the intended hostname. The root domain and
wwware separate names unless one points to the other. - Check the server, firewall, virtual host, and SSL configuration after confirming DNS. A correct DNS response does not prove that the destination is serving the site.
- Compare the answer from the authoritative nameserver with answers from multiple resolvers.
If an A record is not working, first decide whether the returned address is wrong or whether the correct address is returning an application or network error. DNS troubleshooting ends at name resolution; web-server and hosting diagnostics begin afterward. The guide to auditing hosting performance can help with the next layer.
2. CNAME record: a subdomain or SaaS connection fails
- Use the exact target hostname supplied by the service, including spelling and any required trailing dot used by the control panel.
- Enter only the subdomain label in the host field when the interface automatically appends your domain. Accidentally creating
app.example.com.example.comis a common error. - Check whether another record already exists at the same name. A CNAME generally cannot share a name with other ordinary records.
- Confirm that the service also recognizes the hostname in its own dashboard. DNS alone may not complete domain ownership or application setup.
- Check whether the DNS provider uses proxying, flattening, or another routing feature that changes how the record is presented.
For an apex or root domain, follow the DNS provider’s documented method. Some providers offer an alias-style record or flattening instead of a conventional CNAME at the root. Do not substitute an A record for a CNAME unless the service explicitly provides a stable IP address.
3. TXT record: verification or authentication fails
- Confirm the complete string, including punctuation, quotation handling, prefixes, and capitalization where the service requires it.
- Check whether the control panel automatically adds quotation marks. Do not duplicate them unless the interface expects them.
- Look for multiple TXT records at the same hostname. Multiple values can be valid, but combining separate records into one long value can break verification.
- Check that the record is published at the exact requested name, such as the root domain, a selector, or a service-specific subdomain.
- Allow the service to recheck after the authoritative lookup shows the expected value.
TXT records are also used for email authentication. SPF, DKIM, DMARC, and MX records have different purposes and should be checked together rather than treated as interchangeable. See Domain Email and DNS Records Explained for an email-focused checklist.
4. MX record: mail is delayed, rejected, or delivered to the wrong place
- Confirm that MX records exist at the root domain unless the email provider specifies another hostname.
- Check each mail server target and its priority number. Lower preference values are normally tried before higher values.
- Make sure the MX target is a hostname, not an IP address, and that the target itself resolves as required.
- Remove obsolete MX records only after confirming that no active mailbox depends on them.
- Review SPF, DKIM, and DMARC separately. Correct MX records do not by themselves authenticate outgoing mail.
Do not test email changes only from one mailbox or one network. Check the published MX response, then test inbound and outbound delivery while monitoring bounce messages. A domain transfer or nameserver change can affect email even when the website remains online.
5. Nameserver mismatch: edits appear to have no effect
- At the registrar, identify the nameservers currently delegated for the domain.
- Log in to that DNS provider and confirm that the record you edited exists there.
- Compare the zone at the old provider with the zone at the active provider. They may contain different A, CNAME, TXT, or MX values.
- Check for a pending registrar change, expired domain, incorrect nameserver spelling, or DNSSEC configuration that no longer matches.
- After changing delegation, verify the authoritative nameservers before editing individual records again.
Nameserver delegation is the most important branch in the diagnosis. If the registrar points to Provider A but you edit the zone at Provider B, the edit will not be published to the internet. This issue often appears after a domain transfer or a move between hosting platforms.
What to double-check
Use this lookup sequence for nearly any DNS incident:
- Confirm the spelling. Check the domain, hostname, and record type character by character.
- Check delegation. Query or inspect the domain’s nameservers at the registrar and identify the authoritative provider.
- Query the authoritative source. Ask the responsible nameserver for the specific record rather than relying only on a general propagation checker.
- Compare recursive resolvers. Test from more than one network or resolver to distinguish caching from an incorrect published value.
- Inspect related records. Check A and AAAA together, CNAME conflicts, MX targets, and TXT values used by the same service.
- Test the actual service. Open the site, complete verification, send test mail, or use the provider’s validation tool.
TTL indicates how long a resolver may cache an answer; it is not a guaranteed completion time for every network. A high TTL can make an old value persist, while a low TTL does not force every resolver to refresh immediately. Avoid repeatedly changing a record while waiting, because each edit makes the final state harder to identify.
Also check for provider-specific behavior. Some dashboards append the domain automatically, hide the final dot in a target, restrict duplicate records, or place DNS behind a separate product. If a record looks correct in the interface but not in an authoritative query, the interface may be displaying a draft, a different zone, or a transformed value.
Common mistakes
- Editing the wrong DNS zone: the registrar, hosting panel, and DNS provider may all have separate interfaces.
- Duplicating the domain name: a host field containing
www.example.commay becomewww.example.com.example.comwhen the panel appends the domain. - Leaving stale records in place: old A, AAAA, CNAME, or MX values can send different users to different destinations.
- Using a CNAME where another record is required: follow the receiving service’s exact instructions, especially for root domains.
- Changing nameservers without copying essential records: recreate website, email, verification, and security records before or during the delegation change.
- Assuming propagation explains everything: if the authoritative answer is wrong, waiting will not fix it.
- Testing only from a cached device: local DNS caches, browser behavior, VPNs, and corporate resolvers can obscure the current public answer.
Keep a dated export or written inventory of the DNS zone before major hosting changes. This is particularly useful when comparing shared hosting, a VPS, and cloud hosting, where the DNS control point may move independently of the server. For architecture decisions, see Shared Hosting vs VPS vs Cloud Hosting.
When to revisit
Return to this checklist whenever you launch a subdomain, connect a SaaS platform, change web hosting, move email, renew or transfer a domain, or add a verification and security service. Recheck the full zone before seasonal planning cycles, major campaigns, infrastructure changes, and planned migrations rather than waiting for an outage.
Review the workflow when your registrar, DNS provider, hosting control panel, or monitoring tools change. Interfaces and provider features can alter how hostnames, duplicate records, proxying, DNSSEC, and TTL values are handled. A practical recurring review is:
- Export or document the current records and nameserver delegation.
- Mark records that support critical website, application, and email functions.
- Query authoritative answers and compare them with at least one independent resolver.
- Remove confirmed obsolete records and document every intentional exception.
- Test the website, subdomains, verification flows, and inbound and outbound email.
- Keep the worksheet with the hosting and domain renewal records for the next change.
That routine turns DNS troubleshooting from guesswork into a repeatable operational check. When HTTPS is the remaining problem after DNS is correct, continue with the SSL certificate and HTTPS troubleshooting checklist.