How to Check DNS Propagation

Why it isn't instant

When a resolver looks up a record, it caches the answer for the number of seconds set by the record's TTL (time to live). Until that timer expires, the resolver keeps serving the old value even after you've changed it at the authoritative source. Different resolvers cached the record at different times, so they update at different times — which looks like the change "spreading" gradually.

How TTL controls the wait

Plan migrations by lowering the TTL a day or two before the change, making the change, then raising the TTL back afterward. Low TTL during the switch keeps the propagation window short.

• A TTL of 300 (5 minutes) means changes are visible almost everywhere within minutes.
• A TTL of 3600 (1 hour) means up to an hour of stragglers.
• A TTL of 86400 (24 hours) means a full day before every cache expires.

How to check it

Query the record from multiple locations at once and compare the answers. Where they all match the new value, propagation is complete for that record. Where some still show the old value, those resolvers are still inside the old TTL window.

Checking from many geographic locations matters because caches are regional — a record can be fully updated in Europe while North American resolvers still serve the old answer.

Speeding it up (and what you can't)

You can flush your own OS and browser DNS cache, and public resolvers like 8.8.8.8 offer a flush form. But you can't force other people's resolvers to drop their cache — the TTL you set is the real limit. Setting a low TTL before a change is the only reliable way to make the next change fast.