TTLs

Kevin Darcy kcd at daimlerchrysler.com
Mon Jul 16 20:28:27 UTC 2001


Barry Margolin wrote:

> In article <9ivd4b$6ms at pub3.rc.vix.com>,
> Fabiola Caceres  <fabiola at infi.net> wrote:
> >NOW, here is the same zone file, but after I reload my name server:
> >>dig SAO vh4.infi.net
> >
> >; <<>> DiG 8.3 <<>> SAO vh4.infi.net
> >;; res options: init recurs defnam dnsrch
> >;; got answer:
> >;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 4
> >;; flags: qr aa rd ra; QUERY: 1, ANSWER: 0, AUTHORITY: 1, ADDITIONAL: 0
> >;; QUERY SECTION:
> >;;      vh4.infi.net, type = A, class = IN
> >
> >;; AUTHORITY SECTION:
> >vh4.infi.net.           1D IN SOA       ns1.infi.net. hostmaster.infi.net.
> >(
> >                                        2001071400      ; serial
> >                                        3H              ; refresh
> >                                        1H              ; retry
> >                                        1W              ; expiry
> >                                        1D )            ; minimum
> >
> >
> >;; Total query time: 3 msec
> >;; FROM: ns1.infi.net to SERVER: default -- 209.97.61.10
> >;; WHEN: Mon Jul 16 14:32:18 2001
> >;; MSG SIZE  sent: 30  rcvd: 81
> >
> >The only difference I see it's the serial number change, but not the TTL
> >lower to 15 min (900 sec)
>
> The TTL of the SOA record is always copied from the "minimum" field of the
> SOA record.  $TTL is used as the default TTL for all other records.

Actually, the TTL of the SOA record should only be set to MINIMUM when it's being
used for negative-caching responses. The reason that Fabiola got confused, I
think, is because he mispelled SOA as SAO, so the nameserver returned a negative
caching result (because no A record for vh4.infi.net existed). If SOA had been
spelled correctly, then the SOA record returned should have had a 900-second TTL,
as expected.


- Kevin





More information about the bind-users mailing list