Every DNS record has a small number next to it that most people ignore: the TTL. That value influences how long cached answers can remain in use after a website move; it does not guarantee an exact cutover time. Understanding TTL in DNS is one of the easiest ways to make changes faster and safer.
This guide explains TTL in DNS in plain words, how it affects visitors, which values to use, how to plan a migration with it, and the most common mistakes.
TTL (Time To Live) is the number of seconds a DNS resolver is allowed to remember a record before it must ask again. For example:
300 = 5 minutes3600 = 1 hour86400 = 1 daySome DNS panels show options like “Auto,” “1 hour,” or “14400.” Cloudflare’s “Auto” setting is usually 300 seconds for proxied records.

When a visitor’s resolver looks up your domain, it caches the answer for the TTL period. During that time, it won’t check for changes. So:
The TTL in DNS that counts during a change is the one that was active before you edited the record. That’s why changes sometimes seem to take forever — the old answer is still cached. The full story is in DNS Propagation Explained: Why Changes Take Time.
Very low TTLs (like 60 seconds) for everything aren’t necessary for most sites. They don’t make your site faster and can add a tiny delay for visitors whose cache expires more often.
Moving to a bigger server? Use the same planning alongside How to Move From Shared Hosting to a VPS.

Use our dig and nslookup guide to query the relevant name and record type. Keep the selected resolver, timestamp and complete answer with the change record, rather than recording only the displayed TTL.
Keep a written copy of the original record, its TTL and its previous address. This gives you a rollback reference if the new server fails. Check the exact hostname you changed: the root domain and www can use different records, so testing one does not prove that the other is correct.
On a computer with dig installed, run dig example.com A. The number before the record class in the answer shows the remaining cache lifetime returned by that resolver. Repeat the query against another resolver and compare the address as well as the TTL. Replace example.com with your own domain; never paste an unfamiliar address into your DNS panel.
A decreasing value is normal for a cached response. It is not proof that the authoritative record has the wrong setting. If different networks return different addresses, compare their answers with your authoritative nameserver before changing the record again. Repeated edits can make troubleshooting harder.
No. DNS cache duration and page rendering are different concerns. A lower TTL can help resolvers discover future record changes sooner, but it does not optimize images, application code or server response time.
Wait until you have checked the new destination, HTTPS and any related email records. Choose a value suitable for how frequently the record changes and the limits of your DNS provider. Keep the previous host available during the transition whenever possible.
No. Review the individual records involved in your move. Changing an A record does not automatically change the TTL in DNS of an AAAA, MX or TXT record. Record your changes so another administrator can understand the migration later.
For provider-specific settings, read the Cloudflare Time to Live documentation. Your DNS provider may impose a minimum value or manage some records automatically.
Imagine that you manage a small business website and plan to move it from an old hosting account to a new server on Friday evening. The public A record currently has a TTL of 86400 seconds. Someone who queried the record just before you lower its TTL may continue using the previous answer for almost a full day. Saving 300 in the control panel does not send a command to every cache that already holds the old answer.
On Thursday, record the current address and lower the relevant record to 300 if your provider permits it. Leave the address unchanged while the previous cache lifetime passes. Separating these two actions makes it easier to understand what you are testing: first the cache policy, then the destination. If you discover that the original TTL was longer than a day, move the preparation earlier instead of assuming that Thursday is always enough.
Before Friday’s cutover, copy the site, configure the new web server and test the expected hostname there. A DNS change cannot fix an application that is missing its database, a certificate that does not cover the domain, or a firewall that rejects web traffic. Prepare those components while ordinary visitors still use the original host. Keep a final backup and decide who can authorize a rollback.
At the agreed time, synchronize any final content changes and update the A record to the new address. Check whether an AAAA record also points to the previous host. Visitors with IPv6 connectivity may follow that record even when your own computer uses IPv4. Do not remove IPv6 records merely to improve an SEO score; change them according to the actual network configuration of the new host.
After the edit, compare answers from more than one network and test actual pages. Check the home page, a product or contact page and any important form. Watch the old and new server logs if you have access to them. The old server may still receive requests during the transition. Keep it available until you are satisfied that the migration and the remaining cached answers have been handled.

First distinguish a nameserver change from a record edit. Use the DNS record types guide to identify which records support the affected website, email or verification service.
Write down the hostname, record type, existing value, proposed value and current TTL in DNS. Include the root domain and www separately. If the website uses a CNAME, understand the destination of that alias as well. Avoid changing unrelated mail or verification records during a website migration unless the hosting move actually requires it. A smaller change is easier to test and undo.
Some panels accept seconds directly; others offer named choices or an automatic setting. Read the help beside the field before saving. If a provider manages the value automatically, use its documented migration procedure. Do not assume that a number displayed in a tutorial is available in your account, or that a provider’s minimum is the same for every record type.
Write down what a successful move means for the business. For a simple brochure site it may mean that all pages load over HTTPS and the contact form delivers a test message. For a store, it also includes checkout behavior and current inventory. DNS answers alone cannot confirm these application functions. Assign an owner to each check so that a green DNS lookup is not mistaken for a completed migration.
Keep the previous values in a change log outside the DNS editor. If you need to reverse the change, restoring the old address is only one part of recovery: caches that already learned the new address may continue to use it until their lifetime expires. Keep both destinations in a usable state where practical. For a database-backed site, also plan how to handle writes received after the cutover.
A TTL in DNS describes caching of a DNS answer. It is not a timer that starts simultaneously on every visitor’s device when you press Save. Different resolvers fetch the answer at different times, and their remaining lifetimes therefore differ. This is why one colleague may see the new host while another still sees the original one, even when both people test within the same minute.
DNS is also separate from the browser’s cache of page content. If the correct server address is returned but a page still looks old, inspect the website’s cache, any CDN and the application itself. Lowering TTL again will not refresh a cached image or rebuild a page. First identify which layer contains the stale information, then use the controls appropriate to that layer.
Likewise, a nameserver change is a different operation from editing an A record at the current authoritative provider. The delegation that directs resolvers toward your nameservers has its own caching behavior. A short website-record TTL in DNS should not be presented as a promise that switching DNS providers will finish within five minutes. Plan the delegation transition separately and keep the old zone consistent while it remains relevant.
Confirm that you edited the correct DNS provider and the exact hostname. Domain registration, website hosting and authoritative DNS can belong to different companies. A hosting panel may show a DNS zone that is not the one serving your domain. Compare the active nameservers with the provider where you made the change before editing another copy of the record.
Record the answer and its remaining lifetime, then compare it with the authoritative answer and another resolver. If the authoritative data is correct, allow for existing caches instead of repeatedly toggling the address. Keep your notes and the time of the original change. A clear timeline helps distinguish a cached answer from a record that was never updated at the active provider.
Check the destination server’s configuration, certificate and application. Verify that the new host serves the requested hostname rather than a default site. Also check whether a proxy or CDN is part of the request path. DNS can successfully resolve a name even when the service at the resulting destination is unavailable. Treat resolution and application health as separate checks.
Review the mail-related records and the mail service configuration rather than altering the website’s A record again. MX records identify mail destinations, while TXT records can support authentication and verification. Preserve the values supplied by your email provider. If the website and mail moved together, test them independently so that a working home page does not hide an email delivery problem.
A practical log includes the date, the person making the edit, the reason, the previous value and the expected result. Add the old TTL, the new TTL in DNS and the time at which each was saved. This is more helpful than a screenshot without context. If a colleague takes over troubleshooting, they should be able to tell whether a cache preparation step happened before or after the address change.
After the migration, record the checks that passed and any remaining issues. Choose the stable TTL in DNS according to your operational needs, rather than leaving a temporary value in place because nobody remembers why it was set. Keep the migration notes with the hosting documentation. The next person planning a change can use that history to avoid repeating the same investigation.
You can use a short value if your provider allows it and it fits your needs. It is a tradeoff, not an automatic error. Shorter cache lifetimes cause more frequent DNS lookups, while longer lifetimes can reduce that traffic but delay discovery of future changes. Choose deliberately and review the setting when your service becomes stable.
No universal waiting period applies to every record change. The previous cache lifetime, the type of change and the provider’s behavior matter. Use the actual old value and your verification results to plan the transition. A routine record edit and a nameserver migration should not be assigned the same completion promise simply because both involve DNS.
No. A short lifetime does not detect an outage, choose a healthy destination or check that your application works. Monitoring and any failover process need their own design and tests. If you manually change a record during an incident, use the same careful checks and change log that you would use for a planned move.
Prepare before the cutover, keep the previous values, and test the service after the record changes. A lower TTL helps with the caching part of a migration, but it does not remove the need for backups, a working destination or a rollback plan. Treat TTL in DNS as one useful control within a complete website migration process.
Before marking the move complete, ask another person to open the website from a different connection and check a few important pages. Record the result alongside your own tests. A second perspective can reveal that you tested only a local override, an administrator session or a cached page. Keep the test focused on what an ordinary visitor needs to do.
Finally, review who can edit the zone and where the current configuration is documented. Store operational notes in a place your team can find during an incident. TTL in DNS is easier to manage when the next administrator can see the reasoning behind the selected value. Close the migration after the application checks and the DNS checks agree, rather than after an arbitrary countdown.

A lookup result contains several pieces of information that answer different questions. The name identifies the record being returned. The type tells you whether you are reading an address, an alias or another kind of information. The value identifies the destination or data. The TTL in DNS describes the lifetime associated with that answer. Read all of these together rather than checking only that the domain appears somewhere in the output.
For example, an answer for www.example.com does not necessarily describe example.com itself. Similarly, an A response describes an IPv4 address; it does not show whether an AAAA record exists. When collecting evidence, write the complete hostname and record type beside each result. This simple habit prevents a comparison between different records from looking like a propagation problem.
If your tool displays several answers, inspect each one. A service can legitimately use more than one address, and a hostname can resolve through an alias. Do not delete additional answers just because they differ from a simplified tutorial. Compare the configuration with the design supplied by your hosting or DNS provider. If you are unsure why an address is present, investigate its purpose before editing it.
Tell the people responsible for support when the cutover starts, which service is moving and how to report a problem. Give them useful checks: the affected hostname, the page or feature that fails, and whether the issue appears on another network. A report that only says the website is broken leaves too much unknown. Clear observations help you separate DNS symptoms from application errors.
For a business site, schedule changes when someone can remain available to investigate. A short TTL is less valuable if nobody can check the destination after the switch. Leave enough time for a rollback or configuration fix before the next busy period. Avoid combining a hosting move with unrelated design changes, plugin updates and email migrations unless there is a specific operational reason.
Afterward, tell the team what changed and where the final configuration is recorded. Include any temporary measures that still need attention, such as the previous host remaining active. Make each follow-up concrete and assign responsibility. Keeping the old account for a transition period can be useful, but forgetting to review it can leave unnecessary cost and confusion.
The best value depends on how you run the service. A rarely changed record and a record used in an actively managed migration have different needs. Use this guide as a planning framework, then check the provider’s documented limits and your own change process. Consistent records, reliable testing and clear ownership matter more than copying a single number into every field.
Review TTL in DNS after major infrastructure changes, and keep the reason for each setting alongside the record history.
Promotional recommendation: If the DNS change accompanies a website migration, consider KAWEDMA’s website hosting services. Ask for the required destination records, the migration scope and a recovery plan. Confirm responsibility for the authoritative zone and the separate mail checks before changing production records.
We may earn a commission on purchases made through these links, at no extra cost to you.
Don’t miss the chance to participate in the biggest giveaway of this year.