On Apr 16, 2009, at 7:11 AM, <philippe.simo...@swisscom.com> <philippe.simo...@swisscom.com
> wrote:
Hello ISC users
in a special environment, I have to forward 'sub.domain.com', on a
dns where I'm
master for 'domain.com'. all subsequent subdomains of
'sub.domain.com' must
be reachede through this forwarder. I cannot use iteration here
because the
dns authoritative for e.g. 'sub1.sub.comain.com' are not reachable
from this dns.
zone domain.com {
type master;
file "db.domain.com";
}
zone sub.domain.com {
type forward;
forwarders { 1.2.3.4; };
}
all this is is working only if I define a NS for sub.domain.com in
'domain.com' :
----
sub.domain.com IN NS subns.domain.com
subns.domain.com IN A 1.2.3.4
----
I also observed that this NS is not used by bind, because a dummy
address is enough
to do the forward correctly. this NS has only to be present and the
forwarding works.
my question is : is this a bind 'feature' or is this a bad idea to
use that ?
I have to ensure that this behaviour will be the same in the next
bind releases.
What you're seeing is the expected behavior.
- Without the delegation record (the NS record), named thinks it's
authoritative for the zone and won't forward recursive queries.
Instead, it returns a negative answer.
- With the delegation but without the conditional forwarding zone,
named would try to follow the delegation (and glue) in response to a
recursive query.
- With both delegation and conditional forwarding zone, named forwards
recursive queries for the subzone as expected.
The reason to care about the content of the delegation and glue
records is in case the server gets an iterative query, from another
name server. If it will never get such an iterative query, then you
can use bogus data as you have done.
Chris Buxton
Professional Services
Men & Mice
_______________________________________________
bind-users mailing list
bind-users@lists.isc.org
https://lists.isc.org/mailman/listinfo/bind-users
