DNS changes can appear correct in one location while remaining outdated somewhere else. This workflow helps you verify a business website or email change from the authoritative source outward, compare recursive resolver responses, account for TTLs and local caches, and document the result before closing a launch, migration, or infrastructure change.
Overview
“DNS propagation” is commonly used to describe the time it takes for different DNS resolvers, networks, devices, and applications to return a new record. It is not a single switch that updates everywhere simultaneously. A resolver may continue serving a cached answer until that answer’s time to live (TTL) expires, while another resolver has already retrieved the new value.
Before troubleshooting, identify exactly what changed. A website migration may involve an A record, an AAAA record, a CNAME, nameservers, or several records together. An email-provider change may require MX, TXT, and sometimes CNAME records. A domain transfer can introduce a delegation problem if the registrar’s nameserver settings do not match the DNS zone you intended to use.
The most reliable approach is to check in this order:
- Confirm the record at the authoritative nameserver.
- Compare answers from multiple recursive DNS resolvers.
- Check TTLs and allow for cached responses.
- Test from more than one network and device.
- Look for conflicting, duplicated, or incorrectly scoped records.
This distinction matters during a launch. A website can resolve to the new hosting server while some visitors still reach the old server. Email can also behave inconsistently if different senders query different MX or TXT results during a mail migration.
What to track
1. Authoritative nameservers and delegation
Start by checking which nameservers are authoritative for the domain. The registrar’s nameserver settings, the parent zone’s delegation, and the DNS provider’s zone should agree. Query the domain’s NS records using a DNS lookup utility or a reputable online resolver tool. Then query the intended authoritative nameserver directly for the record you changed.
If the authoritative answer is wrong, waiting will not solve the problem. Check for a typo, an edit made in the wrong DNS provider, an unpublished zone, or a domain still delegated to an older provider. During a transfer or DNS migration, confirm that the nameserver set shown at the registrar matches the provider where the records were edited.
2. The changed record and its TTL
Record the hostname, type, value, TTL, and time of the change. For the root domain, the hostname may be displayed as @; for a subdomain, it may be www, mail, or another label. Check both the exact name and the record type. An A record returns an IPv4 address, while an AAAA record returns an IPv6 address. A CNAME points one hostname to another and should not normally be combined with other data at the same name.
TTL is an instruction to caching resolvers, not a guaranteed global deadline. Existing cached answers can remain available until their stored TTL expires, and some systems may behave differently because of negative caching or application-level caching. Avoid repeatedly changing a record while investigating; that makes the expected result harder to identify.
3. Recursive resolver results
After confirming the authoritative answer, compare results from several recursive resolvers. These are the services used by networks and devices to retrieve and cache DNS answers. Query at least one resolver commonly used by your office, one public resolver, and one resolver from a separate network if possible.
Do not treat a propagation-checking website as the authority. It is useful for comparison, but its locations and resolver choices vary. Capture the result, timestamp, resolver address, and response status. A result that differs from the authoritative answer may simply be cached, but it may also indicate that the domain is delegated incorrectly or that a record exists in more than one place.
4. Local, browser, and application caches
If public resolvers show the new value but one computer does not, investigate local caching. Restarting the browser may not clear the operating system’s DNS cache, and clearing the operating system cache may not clear a browser’s internal host cache. A browser can also retain an HTTP redirect, an HSTS decision, or an open connection that makes a DNS change appear ineffective.
Use the DNS cache-flush procedure appropriate to the operating system, then close and reopen the browser. Test with a fresh private window, a different browser, or a command-line lookup. These actions are diagnostic rather than a substitute for fixing the authoritative zone.
Record-change log template
Change date and time (with time zone):
Domain and hostname:
Record type:
Old value:
New value:
TTL before / after:
Authoritative nameserver checked:
Recursive resolvers checked:
Local and alternate-network tests:
Expected behavior:
Observed behavior:
Next checkpoint or owner:
Cadence and checkpoints
Use a checkpoint schedule rather than relying on a single “propagated” message. Immediately after saving a change, verify the provider’s zone and query the authoritative nameserver. At the next checkpoint, compare several recursive resolvers. Then test the actual service from the office network, a mobile connection, and—when practical—a separate location.
For a website launch, check the root domain, the preferred www hostname, HTTP-to-HTTPS behavior, and any application subdomains. Confirm that the destination server presents the expected site and certificate; a correct DNS answer does not prove that hosting or SSL configuration is complete.
For email, query MX records and every relevant authentication record, including SPF, DKIM selectors, and DMARC. A TXT record can contain long or structured values, so compare the complete value rather than a shortened display. Coordinate DNS verification with the mail provider and monitor delivery from more than one sender or mailbox.
For a CDN, failover setup, or website migration, check all active hostnames and both A and AAAA responses. An outdated AAAA record can send IPv6-capable visitors to a different server even when the A record is correct. Keep the old hosting environment available until DNS results and application behavior are consistently correct, unless your rollback plan provides another safe route.
How to interpret changes
| Symptom | Likely causes | First checks |
|---|---|---|
| Stale A result | Cached IPv4 answer, wrong value, or old zone still authoritative | Query the authoritative nameserver, compare TTLs, and confirm delegation |
| Stale AAAA result | IPv6 record was not removed or updated, or it has a separate TTL | Query A and AAAA independently from multiple networks |
| CNAME points to the old service | Alias was edited in the wrong zone or target remains cached | Follow the CNAME chain and inspect each hostname’s authoritative answer |
| MX results differ | Cached mail routing, incomplete migration, or multiple unintended MX records | Compare priority and target values; confirm the intended mail provider |
| TXT verification fails | Typo, split or truncated value, wrong selector, or stale cache | Compare the exact authoritative TXT content with the provider’s requirement |
| Nameserver results are unexpected | Registrar delegation was not changed, has not updated, or points to an old provider | Check registrar settings, parent-zone NS results, and the active DNS zone |
Consistent authoritative results with inconsistent recursive results usually indicate caching or normal resolver timing. Inconsistent authoritative results are more serious: they may indicate multiple authoritative servers are not synchronized, the wrong provider was edited, or the domain is using an unexpected delegation. Escalate to the DNS provider when authoritative servers disagree or return failures.
If DNS looks correct but the website is wrong, check hosting virtual hosts, application redirects, CDN configuration, SSL coverage, and web-server caches. If email is wrong, inspect mailbox routing and authentication separately. DNS is only one layer of a website or email migration.
When to revisit
Return to this workflow after every material DNS edit, domain transfer, email-provider change, CDN setup, hosting migration, or failover test. Recheck sooner if visitors report intermittent access, email delivery changes, certificate warnings, or different results between office and mobile networks.
As an operating practice, review the DNS record inventory monthly or quarterly. Confirm that each record has an owner, a purpose, an expected destination, and a documented TTL. Remove obsolete verification tokens and retired hosting records only after confirming that no service depends on them. Review nameserver access, registrar account security, and renewal contacts as part of the same operational check.
Refresh this article’s tool instructions and platform terminology at least annually, and whenever major DNS providers, operating systems, browsers, or relevant standards change. Lookup interfaces, record limits, cache behavior, and provider workflows can evolve.
Practical closeout checklist: save the change log; verify delegation; query the authoritative record; compare at least two recursive resolvers; test A, AAAA, CNAME, MX, or TXT records as applicable; flush local caches; test from a second network; confirm the website or email service itself; and record the next review date. If all layers agree and the application behaves correctly, close the change with evidence rather than assumption.