DNS propagation is the delay before everyone sees a changed DNS record. Nothing is pushed across the internet: resolvers cache each answer for the record’s TTL (time to live, in seconds) and fetch the new value only when it expires. On WeRegister’s DNS, a change usually goes live within minutes to a few hours.
What is TTL in DNS?
TTL is a number stored with every record. RFC 1034 defines it as “a time limit on how long an RR can be kept in a cache”, where RR means resource record. The administrator of the zone sets it, and a value of zero means the record should not be cached.
A TTL of 3600 lets a resolver reuse the answer for one hour. RFC 2181 allows values from 0 to 2147483647 seconds.
Why is “propagation” really cache expiry?
Your records sit on the authoritative name servers, and they serve the new value as soon as you save it. The delay comes from the many resolvers, run by internet providers, companies and public services, that stored the old answer earlier.
Each resolver refreshes on its own schedule, counted from when it last asked. That is why two people can see different answers for a while.
Missing records are cached too. If someone looks up a name before you create it, the “does not exist” answer is kept for a time taken from the SOA record’s minimum and TTL, so a brand-new record can seem to be missing for a while.
How do you lower the TTL before a planned change?
- Note the current TTL of the record you will change, say 86400 seconds (one day).
- Lower it to a small value, for example 300 seconds, and wait at least as long as the old TTL. By then caches hold the short TTL.
- Make the change, such as pointing the A record to a new server.
- Check the result, then raise the TTL again so resolvers cache less often.
Lowering the TTL does not shorten caches that already hold a long one, so do it ahead of time.
How do you check a DNS record from the command line?
On Windows, macOS and Linux, nslookup works, and dig is available on macOS and Linux:
nslookup -type=A example.eu
nslookup -type=A example.eu 1.1.1.1
dig example.eu A +short
dig example.eu A
dig @ns1.dnshost.example example.eu A
The second command asks a specific resolver, as Microsoft’s nslookup reference shows, which helps when one resolver is stale. A plain dig example.eu A prints the TTL, and repeated queries through a caching resolver typically show it counting down. Asking an authoritative name server, as in the last line, shows the value you saved, free of caches.
Key facts
- TTL is the number of seconds a resolver may cache a record before asking again.
- Propagation is cache expiry: authoritative servers answer with the new value immediately.
- Lower the TTL ahead of a planned change, at least as long as the old TTL, then raise it afterwards.
- A name looked up before it existed can be cached as missing for a while.
- Query an authoritative name server to see what you saved, and a public resolver to see what others see.
Frequently asked questions
How long does a DNS change take? For WeRegister’s DNS, usually from a few minutes to a few hours. The main factor is the TTL of the old record.
Can I force every resolver to update? No. You can lower the TTL beforehand, but not clear other people’s caches.
Is a very low TTL always better? No. It means more queries to your name servers. Keep it low only around a change.
Why do I see the new site but a colleague does not? You use different resolvers, which refreshed at different moments. Both converge once the old TTL has run out.
Manage your records on the DNS and DNSSEC management page. To go further, read how to point a .eu domain at a website and DNS record types explained.