DNS Propagation Checker: How to Diagnose Records That Are Not Updating
DNSDNS TroubleshootingDomain ManagementNetworkingWeb Operations

DNS Propagation Checker: How to Diagnose Records That Are Not Updating

PPlkdt Labs
2026-11-02
7 min read

Use this DNS propagation check to distinguish authoritative record errors, cached answers, delegation issues, and hosting problems.

A DNS propagation check is most useful when it answers a specific question: did the record change reach the authoritative nameservers, or is a resolver, cache, delegation, or configuration mistake showing you an older result? This guide provides a repeatable checklist for diagnosing A, CNAME, TXT, and MX records without relying on a single online checker.

Overview

DNS does not update in one central location. A domain is delegated to authoritative nameservers, which publish the zone records. Recursive resolvers operated by networks, internet service providers, companies, and public DNS services query those nameservers and may temporarily cache the answers. Your device or browser can also retain information locally.

That structure explains why two people can see different answers after a DNS change. It also explains why “propagation” is not always the right diagnosis. If the authoritative nameserver still returns the old value, the change may have been made in the wrong DNS dashboard or zone. If the authoritative answer is correct but one resolver returns the old value, caching or TTL timing is more likely. If every DNS answer is correct but the website still fails, the issue may be hosting, TLS, routing, or application configuration rather than DNS.

Use the following order for a reliable DNS troubleshooting process:

  1. Confirm which nameservers are authoritative for the domain.
  2. Query those nameservers directly.
  3. Compare their answers with several recursive resolvers.
  4. Check the record name, type, value, TTL, and conflicts.
  5. Test the service separately from DNS.

For broader domain, hosting, and email checks, see the DNS not propagating diagnostic checklist.

Checklist by scenario

When an A or AAAA record is not reaching the website

  1. Confirm the intended address. Check the IPv4 address in the A record and the IPv6 address in the AAAA record. Remove an obsolete address only when you are sure no active service depends on it.
  2. Check the hostname. The record for example.com is different from the record for www.example.com. Many DNS control panels display the host as @ for the zone apex, while others expect the full name.
  3. Query the authoritative nameservers. If they return the new address, the DNS change is published at the source. If they return the old address, inspect the provider, zone, and record name before waiting.
  4. Test the server directly. An address can resolve correctly while the web server rejects the hostname, serves the wrong virtual host, blocks traffic, or lacks a valid certificate.

If this follows a deployment, separate DNS verification from application verification. A deployment guide for a static site or a server using Nginx can help isolate those layers: custom-domain static site deployment and Node.js on a VPS with Nginx.

When a CNAME record is not working

  1. Verify that the CNAME is attached to the exact subdomain being tested.
  2. Check whether the destination should be a hostname, not an IP address. CNAME records point to names; A and AAAA records point to addresses.
  3. Look for an existing A, AAAA, or other conflicting record at the same hostname. A CNAME generally cannot coexist with ordinary data at that name.
  4. Check whether the hosting platform requires a separate domain verification record or an exact target hostname.
  5. Test the destination independently. A correct CNAME does not guarantee that the destination service is configured to accept your custom hostname.

When a TXT record is not being detected

TXT records are commonly used for domain verification, SPF, DKIM, DMARC, and other configuration checks. Confirm that the record was added to the correct host. SPF normally belongs at the domain root, while DKIM usually uses a selector-based hostname such as selector._domainkey. DMARC normally uses _dmarc.

Check quotation marks and long values carefully. Some dashboards split long TXT strings into multiple displayed segments; that can be valid if the resulting TXT data matches the required value. Avoid adding multiple SPF records at the same hostname. If you are troubleshooting email authentication, review the related DNS automation and record management guide before changing records programmatically.

When email delivery fails after an MX change

  1. Confirm the MX record is at the domain used in recipient addresses.
  2. Check the MX priority values. Lower preference numbers are generally tried before higher numbers.
  3. Verify that each MX target is a hostname with a valid A or AAAA record; do not use an IP address directly as an MX target.
  4. Remove obsolete MX records only after confirming the intended mail provider and migration state.
  5. Check SPF, DKIM, and DMARC separately. Correct MX records route mail, but they do not by themselves authenticate outgoing messages.

What to double-check

Authoritative nameserver delegation

Find the nameservers listed for the domain at the registrar or with a command-line lookup. Then query each authoritative nameserver directly. All authoritative servers should normally provide a consistent answer. A mismatch can indicate that the zone was updated in one place but not another, or that the domain is delegated to a different provider than expected.

TTL and cached answers

TTL, or time to live, tells recursive resolvers how long an answer may be cached. It is not a guaranteed countdown for every device or service, and lowering TTL before a change does not instantly remove values already cached under a previous TTL. After the authoritative answer changes, some resolvers may continue returning the earlier value until their cache expires.

Compare results from your normal resolver and at least one independent resolver. A difference is useful evidence, but it is not proof that the entire internet is seeing one result. Corporate DNS, VPN software, security filters, operating systems, browsers, and home routers can all affect what you observe.

Record syntax and provider conventions

Review the record type, host, value, and trailing dot behavior. Some control panels automatically append the domain name; entering a fully qualified name where a relative name is expected can produce an unintended hostname. Also check whether a proxy, flattening feature, or provider-specific alias record changes how the answer is presented.

Delegation and DNSSEC

If nameserver delegation is wrong, records in the intended DNS dashboard will not affect the public domain. If DNSSEC is enabled and its delegation data is stale or incorrect, validating resolvers may report a failure even when the zone appears correct at the authoritative provider. Treat delegation and validation errors as separate checks from ordinary record propagation.

Common mistakes

  • Checking only one propagation website. Aggregated tools are convenient, but direct queries reveal whether the issue is authoritative, recursive, or local.
  • Editing DNS at the wrong provider. The registrar, web host, CDN, and DNS host may be different companies. The authoritative nameservers identify where the active zone is managed.
  • Forgetting the www hostname. Updating the apex record does not automatically update www, and the reverse is also true.
  • Adding duplicate records. Duplicate A, MX, SPF, or verification records can create inconsistent service behavior.
  • Assuming DNS controls the whole request. A successful lookup does not confirm firewall rules, ports, web-server routing, certificates, redirects, or application health.
  • Changing several records at once. Make one controlled change where possible, record the previous value, and test after each change.
  • Ignoring automation drift. Terraform, a Cloudflare API script, or a deployment pipeline may restore an old value after a manual edit. Check the source of truth before repeating the change.

For deployments that combine Docker, GitHub Actions, and Nginx, use a layered test plan rather than treating a failed URL as a DNS-only problem. The Docker, GitHub Actions, and Nginx deployment guide provides a useful related workflow.

When to revisit

Return to this checklist whenever a domain changes hosting, CDN, registrar, email provider, certificate setup, or DNS automation. It is also worth reviewing before seasonal traffic changes, planned migrations, or updates to deployment workflows. Those events can alter the authoritative provider, record set, TTL strategy, server address, or automation source of truth.

Before making a change, capture the current authoritative answers and export or record the relevant zone values. During the change, note the time, intended record, previous value, and expected TTL. Afterward, run this short sequence:

  1. Query the authoritative nameservers.
  2. Compare at least two recursive resolver results.
  3. Flush or restart only local caches that you control, then test again.
  4. Verify the service using the exact hostname, including www or another subdomain when relevant.
  5. Check email, verification, or API clients separately if the change affects them.
  6. Document the final record and remove temporary migration entries when they are no longer needed.

If the authoritative record is still wrong, fix the DNS zone or delegation. If it is correct but external answers differ, allow for caching and monitor resolver results. If DNS is correct and the service still fails, move on to hosting, TLS, firewall, reverse proxy, or application logs. This separation turns a vague “DNS not propagating” report into a short, repeatable diagnostic process.

Related Topics

#DNS#DNS Troubleshooting#Domain Management#Networking#Web Operations
P

Plkdt Labs

Developer Tools and Infrastructure 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.