[dns-operations] TTL analysis: "How Long Does a Second Actually Last?"

Jared Mauch jared at puck.nether.net
Fri Mar 10 17:15:26 UTC 2017

On Fri, Mar 10, 2017 at 10:52:28AM -0500, David C Lawrence wrote:
> Interesting but five year old page I'd only just now come across for
> the first time.  Thought some of you would find it interesting as
> well.  Apologies if it has actually been discussed before and I just
> missed it.
> https://00f.net/2011/11/17/how-long-does-a-dns-ttl-last/

	Few other interesting behaviours out there:

	1) unbound has serve-expired, which other dns implemenations
have, this minimizes impact when authorities are down or unreachable.

	2) should look at the "zombie" research as well, this relates
to TTL in a related work.  Many qnames seem to live on for extended
periods of time:


	3) When using an unnamed vendors load balancer for DNS queries
back in the early 90s, it had serious issues with TTL expiry handling
when the replies were sub-10 second TTLs.

	- Jared

Jared Mauch  | pgp key available via finger from jared at puck.nether.net
clue++;      | http://puck.nether.net/~jared/  My statements are only mine.

More information about the dns-operations mailing list