IPv4 not working reverse on > /24 cidr

[Ryan Pavely]

   Ryan Pavely
    Net Access Corporation
    http://www.nac.net/

On 7/22/2013 11:00 AM, Barry Margolin wrote:
> In article <mailman.877.1374504592.20661.bind-users at lists.isc.org>,
>   Ryan Pavely <paradox at nac.net> wrote:
>
>> Ok.  What am I doing wrong?  As far as I know this has worked for years
>> and sometime, weeks, months, years, ago it stopped.
>>
>> This is for doing > /24 (greater in cidr smaller in size)
>> Example: we have a /25 that we host... and another /25 we host.. so we
>> split it up into smaller files unless we own the entire/24
>>
>>
>> The config is loaded.
>> Rndc reload reports all is well.
>> But a lookup fails.
>>
>> Help?
>>
>>
>> BIND 9.9.3-P1 on Linux
>>
>> == included file in named.conf
>> zone "128/27.1.10.10.IN-ADDR.ARPA" {
>>           type master;
>>           file "/usr/named/rev/10.10.1.128.rev";
>> };
> Do you also have a 1.10.10.in-addr.arpa zone, and does it have all the
> necessary CNAME records pointing x.1.10.10.in-addr.arpa to
> x.128/27.1.10.10.in-addr.arpa?
>

I do not.  10.10.1.128/27 is a RFC1918 sample.  In a real-world example 
I also have some ATT address space 12.44.51.192/27 or so.. They point it 
to me.

If I host a partial class, in this case 10.10.x.x I need to have a 
parent file that cnames?
     Am I correct I would do something like the following...


     $GENERATE 128-160 $ CNAME $.128/27.1.10.10.IN-ADDR.ARPA.



What about when the block is already cnamed -> pointed -> delegated to 
my host from an external source?

     I tested this.  It appears to be true.  Interesting.


So that would suggest any time any block > a /24 is hosted you must 
actually host the parent zone, pointing to the larger cidr, and then 
have your normal files for each cider in that block.