Re: Build BIND 9.9.3-P1 on Solaris 10 with 'cc', using OpenSSL built with 'gcc'?

2013-06-10 Thread mikep

On Sun, 9 Jun 2013, Dennis Clarke wrote:


That seems oddthough I haven't tried building 9.9.3-P1 yet.

But, all the previous releases built with gcc.  Our Solaris package
build/management system only has gcc.

BIND 9.9.3 was the first BIND that got built 64-bit, which did take a
little extra work in getting it find our 64-bit builds of openssl and
zlib.

Which was basically to have it look in /usr/local/lib/(amd64|sparcv9)
instead of just /usr/local/lib (had found in config.log that it was
complain about architecture mismatch.)


I have been building bind 9.x for years on Solaris with cc and no problem.
As for OpenSSL, well the issue seems trivial, most likely there is a downstream
lib dependency on libgcc.so or some gcc c++ libs perhaps. One merely needs
to setup a LD_OPTIONS correctly with care paid to LD_RUN_PATH and you
should be fine.

Ultimately Oracle Studio 12.3 ( or Sun Studio or Forte or whatever the brand
name du jour is ) doesn't care where you put your libs so long as you say
where to find them and have a match on the architecture.


Although I have 'crle' set to include the GNU/GCC libraries,
adding "LDFLAGS='-L/opt/gnu/lib -R/opt/gnu/lib'" to the 'configure'
command has got it all working with 'CC=cc' -- thanks for the pointer!


Dennis


Mike
--
Mike PetersonInformation Security Analyst - Audit
E-mail: mi...@noc.utoronto.caWWW: http://www.noc.utoronto.ca/
Tel: 416-978-5230   Fax: 416-978-6620
___
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


Stub zones vs minimal responses

2013-06-10 Thread Chris Buxton
I'm seeing something I didn't expect in BIND's behavior, and I wanted to get 
confirmation from someone that this is expected, or at least a known limitation.

If an authoritative server is configured to send minimal responses, will a stub 
zone get all the necessary data from that server? What I'm seeing is, the 
recursive server sends an SOA query; the response contains only the SOA record, 
and no NS or A records. The recursive server doesn't follow up with an NS 
record query, and therefore the stub zone fails. Queries to the recursive 
server for data in that zone get a SERVFAIL response.

Am I understanding the evidence correctly?

Regards,
Chris Buxton
___
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