Skip to content
SpoofKiller
Free tool · no signup

You saved the record. Has anyone else noticed?

Ask several major public DNS resolvers the same question at the same moment and compare the answers. If they agree, your change has reached them. If they do not, you can see who is still holding the old record, and for how long.

What the checker does

Six independent caches, one question, side by side.

Works for A, AAAA, MX, TXT, CNAME, and NS records, including underscore names like _dmarc.example.com and selector._domainkey.example.com.

Parallel queries
Cloudflare, Google Public DNS, Alibaba Cloud DNS, DNS.SB, NextDNS, and AdGuard are each asked over DNS-over-HTTPS, with a four-second limit apiece. A slow resolver does not hold up the rest.
The majority answer
The record set most resolvers return, and a mark against any resolver whose answer differs. Record order is ignored; content is not.
TTL remaining
How many seconds each resolver will keep its cached copy before asking your nameservers again. This is the honest answer to “how much longer?”
Response time and DNSSEC
How quickly each resolver replied, and whether it marked the answer as DNSSEC-validated.
Failures, labeled as failures
A resolver that times out or errors is shown as not answering. It is not counted as “record missing,” because it is not evidence of anything about your record.
What it is not
These are resolver operators, not locations on a map. Each uses anycast, so which of its sites answers depends on where our server sits. The tool compares caches; it does not survey the planet.
How to read the result

Nothing propagates. Caches just expire.

“DNS propagation” suggests your new record is traveling outward, like news. It is not. Your nameservers have the new record the moment you save it. What takes time is every resolver on the internet finishing with its cached copy of the old one. Each copy carries a TTL, a countdown set by you, and when it reaches zero the resolver asks again and gets the new answer.

So the question is never whether the record has spread. It is which caches have expired. When all six resolvers here agree on the new value, the large public ones are done, and the rest of the internet will follow within the old TTL. When they disagree, you are watching the changeover happen. When they all return the old value hours later, the change probably did not save, or was saved somewhere other than the nameservers the domain actually uses.

For email records, the most common false alarm is a name entered twice. A DMARC record belongs at _dmarc.example.com; many DNS editors add the domain for you, and typing it in full produces _dmarc.example.com.example.com, which resolves nowhere useful. If the checker finds nothing, look there first. Once the record is visible, validate what it says: the DMARC checker, SPF checker, and DKIM checker each read the content, not just its presence. New to publishing these? Start with how to configure DMARC. And when you would like to be told about changes you did not make, that is domain security monitoring.

Two hands patching cables into a switchboard whose lamps glow green, amber, and red.
Every resolver keeps its own copy of the directory, and updates it on its own schedule.
Common problems and fixes

When the answers are not what you expected.

  • Every resolver says the record does not exist

    Check the host field at your DNS provider for a doubled domain, and check that you edited the zone at the provider your NS records point to. Query NS for the domain here to confirm which that is.

  • Some resolvers show the old value

    Normal during a change. Note the largest “TTL left” figure: that is the longest you should have to wait. Next time, lower the TTL a day before a planned change.

  • Resolvers disagree for more than a day

    Your nameservers may be serving different data, which happens when a secondary has not synchronized. Compare the answers from each authoritative nameserver at your DNS host.

  • A TXT record appears twice

    Two TXT records at one name are both returned. For SPF and DMARC that is an error, not redundancy: delete the one you did not mean to keep.

  • The CNAME shows, and the TXT lookup is empty

    A name with a CNAME cannot hold other records. The TXT content has to exist at the CNAME's target, which for delegated DKIM selectors is your provider's responsibility.

  • The record is visible and mail still fails

    Visible is not the same as correct. Run the domain health check to validate the content of every email record together.

How to

How to check DNS propagation

Confirm that a new or changed record is visible to major public resolvers.

  1. 1

    Enter the full name

    Type the exact hostname of the record, such as _dmarc.example.com, and choose the record type.

  2. 2

    Compare the answers

    Each resolver's answer is shown with a mark for whether it matches the majority.

  3. 3

    Read the TTL

    For any resolver showing the old value, “TTL left” is how long until it asks your nameservers again.

  4. 4

    Validate the content

    Once the record is visible, run it through the matching checker to confirm it says what you intended.

See all twelve free tools

FAQ

Questions about this tool

Straight answers. If yours is not here, ask us and a person will reply.

Ask a question

Free, no obligation

A record that will not show up?

Send us the domain and the record you tried to publish. We will look at where it actually landed and tell you what to change.

  • A person reads every message
  • No sales sequence, no list rental
  • Reply within one business day

We use your details to reply. Nothing else.

You published it. Now find out what it changed.

Create a free account to collect the DMARC reports your new record asks for, and read them as a dashboard.