It’s always DNSHow long does DNS propagation take?
Nothing is pushed anywhere when you change a DNS record. Every resolver that already has the old answer keeps serving it until its cached copy expires. “Propagation” is just those caches running out, one by one.
The short answer
| What you changed | Worst case | Set by |
|---|---|---|
| A, AAAA, CNAME, MX, TXT record | The record’s old TTL | You (your DNS provider) |
| A name that didn’t exist before | The zone’s negative TTL (often 5 to 60 minutes) | Your zone’s SOA record |
| Nameservers (NS) at the registrar | Up to 48 hours for .com and .net | The TLD registry |
Why records follow their TTL
Every DNS answer carries a TTL: the number of seconds a resolver may reuse it. If your A record had a TTL of 3600 when Google’s resolver last fetched it, Google can keep serving the old address for up to an hour after you change it. Big resolvers run many separate caches, so each cache expires on its own schedule. That’s why one lookup can show the new answer while the next still shows the old one.
Why “doesn’t exist” sticks around
If someone looked up a name before you created it, resolvers cached the “doesn’t exist” (NXDOMAIN) answer. That negative answer is kept for the smaller of the SOA record’s TTL and its minimum field. For names under .com that come back NXDOMAIN from the .com servers, that is 900 seconds (15 minutes) per name.
Why nameserver changes are slow
When you change nameservers, the list lives in the TLD’s zone, not yours. The .com and .net registry publishes delegations with a TTL of 172,800 seconds, so resolvers may keep using your old nameservers for up to two days. You can’t shorten this. Keep the old nameservers serving correct answers until the window has passed.
How to make a change land faster
- Lower the TTL first. A day before the change, drop the record’s TTL to 60 or 300 seconds, then wait for the old TTL to expire before switching.
- Purge the public resolvers. Google, Cloudflare and OpenDNS let anyone ask them to drop a cached record. See how to flush DNS caches.
- Clear the caches close to you. Your OS, browser and any internal resolver (Kubernetes CoreDNS, a company DNS server, AWS’s VPC resolver) keep their own copies.
- Watch it happen. The checker on this site shows each resolver’s answer and a live countdown to when its copy expires.
Resolvers that bend the rules
Some resolvers cap TTLs or keep answers longer than asked. AWS’s VPC resolver, for example, never reports a TTL above 300 seconds to instances, and many resolvers will serve a stale answer if your nameservers stop responding. A lookup tool that only queries public resolvers can’t see these internal caches.
See what Google, Cloudflare, OpenDNS, Quad9 and major ISPs are serving, how long each will cache it, and the registry status.
Open the checkerMore guides: How to flush DNS caches · clientHold and serverHold