[dns-operations] Prime TTL values for TLD and root server delegations.

[Jim Reid]

On 21 Dec 2009, at 18:24, Colm MacCárthaigh wrote:

> Having the NS record and the A record associated with that NS expire  
> at the same
> is sub-optimal.

In your opinion. :-)

> What's more, the behaviour will affect the most frequently queried  
> records. I agree that the problem happens rarely, but it's still  
> avoidable.

In other words, this is at best a micro-optimisation that gets lost in  
the noise. Assuming it could be detected or even measured.

Some back of the envelope numbers might help here. Assume a resolving  
server speaks to 50 TLDs, each of which has 2 day TTLs on their NS  
RRsets and accompanying A/AAAA records. That means on average around 1  
TLD "refresh" per hour. That server is probably getting tens of  
queries per second from local clients. So you're looking to optimise  
something which might affect 0.003% of those local queries by adding  
perhaps a 50-100ms latency to the query that gets the cache miss.