[Pdns-users] Re: Some PowerDNS Recursor oddities

[bert hubert]

On Tue, May 16, 2006 at 03:00:27PM +0200, Stephane Bortzmeyer wrote:
> > You may get a warm feeling to be compliant with the cited paragraphs
> > from 2181 but you are doing your users a disservice in real life.
> 
> Care to elaborate? Because caching doubtful data instead of
> authoritative data does not seem to help the users :-)

This is only about the TTL, nothing else. Lots of domains have very low TTLs
which means a lot of additional queries to the registry, which did not ask
for this. These additional queries slow things down a lot.

You think this helps users? Also, do you consider the registry doubtful?
Do you crave additional queries?

Also, what does it *mean* to have a lower TTL than specified in the parent
zone? It means that either:

1) the parent zone is making false statements about the expected
validity of its answers 
2) You are making changes faster than the registry can follow in any case

Either of these cases is broken. Should I make a change to facilitate these
broken assumptions and slow down the DNS process for everybody? And burden
the registry with more queries?

I don't see it.

If you really want this, the feature will be called 'auth-ttl-override'. I
wonder how many people will turn it on.

But perhaps my understanding of the implications and uses of mismatched ttls
is broken. I'd love to hear.

Kind regards,

bert hubert.

-- 
http://www.PowerDNS.com      Open source, database driven DNS Software 
http://netherlabs.nl              Open and Closed source services