Why would you use forwarding over links that are "neither fat nor reliable"? Are you a masochist? Replication of the data is much recommended over such links...

As for your "pecking order", what distinction are you drawing between forwarding and recursion? Forwarding is recursive. The high-level distinction is between having the data authoritative locally and not having it authoritative locally. If you want to make a finer distinction within the not-locally-authoritative case, then make the distinction between recursive (e.g. forwarding) and iterative (e.g. stub, or delegation from an internal root zone).

                                    - Kevin
On 6/3/2013 2:53 PM, Alan Shackelford wrote:

I agree with Len. Whenever we merge a new location into our network, and the circuit is neither fat nor reliable, I make their DNS forward queries for our internal zones to us, keep authority for their own zones, and do recursion for everything else. This allows us to serve the users as we slowly homogenize the whole mess. The "pecking order" seems to be authoritative first, forward second, and recursion last.

Alan

Alan V. Shackelford Senior Systems Software Engineer

The Johns Hopkins University and Johns Hopkins Medical Institutions

Baltimore, Maryland USA ashac...@jhmi.edu <mailto:ashac...@jhmi.edu> 410-735-4773

*From:*bind-users-bounces+ashackel=jhmi....@lists.isc.org [mailto:bind-users-bounces+ashackel=jhmi....@lists.isc.org] *On Behalf Of *Leonard Mills
*Sent:* Sunday, June 02, 2013 3:29 PM
*To:* Jonathan Reed; bind-users
*Subject:* Re: does zone trump forward?

As I understand AUTHORITATIVE trumps anything. For example, from an inside intranet name server forward the root (".") to somewhere on your edge, sprinkle in a few internal-only authoritative zones, and enjoy. This is certainly not the only choice, but it functions pretty well.

Len

    ------------------------------------------------------------------------

    *From:*Jonathan Reed <cronst...@gmail.com
    <mailto:cronst...@gmail.com>>
    *To:* bind-users <bind-users@lists.isc.org
    <mailto:bind-users@lists.isc.org>>
    *Sent:* Sunday, June 2, 2013 12:10 PM
    *Subject:* does zone trump forward?



    I've only ever come across bind configs where forwarding is in
    place to locate certain zones, then all other queries are handled
    by either recursion or authoritatively. But what about the other
    way around, where I'm master for a few zones but forward the rest?
    Consider this:

    view "the-internet" {

      recursion no;

      type forward;

      forwarders { 8.8.8.8; };

      zone "example.com <http://example.com/>" {

          type master

          file "example.com <http://example.com/>"

      ......

    }

    Whats confusing me is the implied configuration setting of forward
    first when the forward statement is used. If it truly forwards
    first, then I see an odd logical scenario happening. All queries
    are sent to the forwarder before being handled by localhost. Then,
    once the forwarder recognizes that I'm the master of example.com
    <http://example.com/>, why would a loop not occur if the forwarder
    matches this view?

    To ask the question another way, does the zone statement take
    precedence on matching queries over any forwarding?

    Thanks


    _______________________________________________
    Please visit https://lists.isc.org/mailman/listinfo/bind-users to
    unsubscribe from this list

    bind-users mailing list
    bind-users@lists.isc.org <mailto:bind-users@lists.isc.org>
    https://lists.isc.org/mailman/listinfo/bind-users



_______________________________________________
Please visit https://lists.isc.org/mailman/listinfo/bind-users to unsubscribe 
from this list

bind-users mailing list
bind-users@lists.isc.org
https://lists.isc.org/mailman/listinfo/bind-users

_______________________________________________
Please visit https://lists.isc.org/mailman/listinfo/bind-users to unsubscribe 
from this list

bind-users mailing list
bind-users@lists.isc.org
https://lists.isc.org/mailman/listinfo/bind-users

Reply via email to