DNS Propagation Checker

Query a record against 8 major public DNS resolvers at once to see whether a recent change has caught up everywhere yet. Needs a small server-side check, since browsers can't query a specific external resolver directly.

Enter a domain and click "Check propagation".

Why resolvers disagree right after a change

Every public resolver keeps its own independent cache of DNS answers, and each cached answer stays valid for exactly as long as its TTL (time to live) says. When you change a record, any resolver that already had the old answer cached keeps serving it until that TTL runs out — regardless of when you made the change. That's the entire mechanism behind "DNS propagation": there's no actual push notification telling resolvers a record changed, just independent caches expiring on their own separate schedules.

This means propagation time is bounded by your old record's TTL, not your new one. If you're planning a DNS change and want it to take effect quickly everywhere, lower the TTL a day or two beforehand — once the low-TTL version has had time to fully propagate, the actual change will clear out of every resolver's cache much faster.

What "8 resolvers" actually means here

This checks against 8 major public resolver providers — Google, Cloudflare, Quad9, and others — each queried directly on their own IP address using the raw DNS protocol. These are large anycast networks rather than single physical machines in one place, so there isn't a meaningful "city" to label each result with the way some tools imply; what you're really seeing is whether different major resolver operators have caught up with your change, which is what actually matters for "can people reach my new setup yet."

Frequently asked questions

Why do different DNS resolvers show different answers for the same domain?

Each resolver caches answers independently until its own TTL expires — right after a change, some will still be serving the old cached answer.

How long does DNS propagation actually take?

It's bounded by your previous record's TTL, not the new one. Lowering TTL ahead of a planned change shortens how long the old answer can linger afterward.

Does this check DNS servers in different physical locations?

It checks different major resolver providers, each on their own IP — most run as anycast networks without one true physical location, so results are labeled by provider, not city.