DNS Propagation Troubleshooting Guide: Check Nameservers, TTLs, Caches, and Recent Changes
DNSDNS propagationnameserversTTLDNS troubleshootingwebsite operations

DNS Propagation Troubleshooting Guide: Check Nameservers, TTLs, Caches, and Recent Changes

wwhata.cloud Editorial Team
2026-11-21
7 min read

A practical DNS propagation troubleshooting checklist for nameservers, TTLs, caches, records, websites, subdomains, email, and DNSSEC.

DNS changes can appear broken even when the record is correct. This reusable DNS propagation troubleshooting guide helps you separate nameserver errors, authoritative DNS problems, resolver caches, TTL timing, and incorrect records before you change anything else.

Overview

When a DNS change is not working, “propagation” is only one possible explanation. A domain may still point to the old service because the registrar delegates it to different nameservers, the authoritative zone was edited in the wrong control panel, a recursive resolver is serving a cached answer, or the record itself conflicts with another record.

DNS troubleshooting becomes much faster when you compare two answers: the answer from the domain’s authoritative nameserver and the answer returned by a public or local recursive resolver. The authoritative answer shows what the DNS provider is publishing. The recursive answer shows what a user, office network, ISP, or application may still have cached.

Before starting, record the intended change. Use this inventory for every website launch, hosting migration, subdomain, or email-provider update:

  • Domain and hostname: for example, example.com or app.example.com.
  • Record type: A, AAAA, CNAME, MX, TXT, or another type.
  • Previous value and intended value: include the complete IP address, target hostname, or text value.
  • Nameservers in use: copy the delegated nameservers shown at the registrar.
  • Change time and TTL: note when the edit was saved and the TTL visible in the response.
  • Expected service: website, API, email, verification, or another function.

For background on record types, use the DNS records cheat sheet and zone-file template. If you are moving infrastructure, also keep a copy of the cloud server migration runbook beside this checklist.

Checklist by scenario

1. A website or root domain still opens the old server

  1. Check the domain’s delegated nameservers at the registrar. A record changed at Provider A will not affect visitors if the domain is delegated to Provider B.
  2. Query the authoritative nameserver directly. Replace the example nameserver and domain in this command:
dig @ns1.example-dns.com example.com A +noall +answer
  1. Query several recursive resolvers for comparison:
dig @1.1.1.1 example.com A +noall +answer
dig @8.8.8.8 example.com A +noall +answer
dig @9.9.9.9 example.com A +noall +answer
  1. Compare the returned IP address with the intended server. If the authoritative answer is wrong, this is not a propagation problem: inspect the zone, record name, record type, and active nameservers.
  2. If the authoritative answer is correct but recursive answers differ, wait for cached responses to expire according to their remaining TTL. Test again from the affected network and from an independent connection.

2. A new subdomain does not resolve

Confirm that the record name was entered as the provider expects. Some DNS panels require only the label, such as app; others accept the full hostname. Check whether the intended record is an A record pointing to an IP address or a CNAME pointing to a hostname. A CNAME should not normally be combined with other data at the same name, and a CNAME target must itself resolve.

Test both the exact hostname and the parent domain:

dig app.example.com A +noall +answer
dig app.example.com CNAME +noall +answer
dig example.com NS +noall +answer

A response of NXDOMAIN indicates that the name does not exist from that resolver’s perspective. An empty answer with an authority section can mean the name exists but has no record of the requested type. These cases require different fixes, so save the complete response when escalating.

3. Email stopped working after a DNS change

Inspect MX records first. The MX target should be a hostname, not an IP address, and that target must resolve. Then review related TXT records used for SPF, DKIM, and DMARC. A website record can be correct while email remains broken because mail routing and authentication use separate records.

dig example.com MX +noall +answer
dig example.com TXT +noall +answer
dig _dmarc.example.com TXT +noall +answer

Do not delete existing email records simply because a provider supplied new values. Compare the old and new configuration, check whether multiple services are still required, and verify the provider’s exact formatting. For a fuller audit, see the SPF, DKIM, and DMARC setup guide.

4. Nameservers were recently changed

Nameserver changes affect delegation, not just an individual record. First confirm the registrar displays the intended nameservers without spelling errors. Next query the parent delegation:

dig example.com NS +trace

Use the trace to identify where delegation leads. Then query each new authoritative nameserver directly and confirm that the complete zone exists there, including website, subdomain, and email records. A common failure during DNS migration is creating only the new A record while omitting MX, TXT, verification, or DNSSEC-related configuration.

5. DNSSEC was enabled or changed

If some resolvers report a DNSSEC validation error while others appear normal, inspect the DS record at the parent and the signing keys at the authoritative provider. A stale DS record, mismatched key, or incomplete rollover can make a valid-looking zone fail validation. Use your DNS provider’s DNSSEC status tools and consult the DNSSEC setup and troubleshooting guide before removing security settings as a first response.

What to double-check

TTL affects how long recursive resolvers may retain an answer, but it does not force every cache to refresh at the same moment. A practical reference is:

SituationWhat to expectWhat to check
Authoritative record changedThe DNS provider should show the new value quickly after saving.Correct zone, hostname, type, and nameserver.
Recursive answer is oldThe old value may remain until its cached TTL expires.Remaining TTL and several independent resolvers.
Nameservers changedDifferent networks may observe old or new delegation during the transition.Registrar delegation, parent NS response, and complete new zone.
Negative resultA resolver may cache a missing-name response for a period set by the zone.NXDOMAIN versus empty answer and the SOA response.

Test from at least three perspectives: the authoritative server, a public recursive resolver, and the affected local network. Browser behavior is not a reliable DNS test because browsers, operating systems, routers, VPNs, and security software may retain their own caches. Flush a local cache only after recording the result; flushing cannot repair an incorrect authoritative record.

Also check for IPv6. A correct A record does not help a client that receives an incorrect AAAA record and prefers IPv6. For a website, compare both:

dig example.com A +noall +answer
dig example.com AAAA +noall +answer

Finally, DNS success does not guarantee application success. Once the hostname resolves, test the server, firewall, virtual-host configuration, TLS certificate, redirects, and application health. DNS answers only establish where a client should try to connect.

Common mistakes

  • Editing the wrong DNS provider: the registrar, hosting panel, and authoritative DNS provider may be separate accounts.
  • Using the wrong record name: an automatically appended domain can turn an intended hostname into a different one.
  • Replacing rather than reviewing records: a migration can unintentionally remove MX, TXT, DKIM, verification, or service records.
  • Assuming every delay is propagation: a typo, wrong IP, stale delegation, or broken target will not resolve with time.
  • Testing only from one network: one office resolver may be cached while another network has the current answer.
  • Ignoring the zone apex: many providers treat the root domain and www as separate names requiring separate configuration.
  • Changing records repeatedly: each edit makes the evidence harder to interpret and may reset the period during which caches hold different answers.

When to revisit

Use this checklist whenever you change nameservers, move a website to cloud or VPS hosting, launch a subdomain, switch email providers, configure SSL validation, or alter DNSSEC. Revisit it before a planned launch and after the change has had time to reach recursive resolvers. During seasonal planning cycles, review the domain inventory, renewal ownership, DNS access, backup contacts, and documented TTLs rather than waiting for an outage.

Update your internal runbook when your registrar, DNS provider, hosting panel, monitoring tools, or deployment workflow changes. Provider interfaces and resolver-testing methods evolve, so keep the exact commands, active nameservers, escalation contacts, and rollback values current. For a broader launch review, use the website launch checklist. If you are changing registrars, follow the domain transfer checklist and verify DNS before and after the transfer.

Final action: before declaring a DNS change complete, save the change inventory, authoritative responses, recursive comparisons, and a real service test. If the authoritative answer is wrong, fix the zone. If it is right but caches differ, monitor TTL expiration. If DNS is correct but the service fails, move on to hosting, networking, TLS, or application diagnostics instead of making more DNS edits.

Related Topics

#DNS#DNS propagation#nameservers#TTL#DNS troubleshooting#website operations
w

whata.cloud Editorial Team

Technical Editor

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.