[Samba] dynamic DNS Updates still failing, re-installed 9 more times, tried everything I could think of, now bald.
garymaurizi at gmail.com
Sun Jun 2 21:16:00 MDT 2013
I think I might have figured out something about this Centos 6.4 thing and
BIND9_DLZ dynamic updates NOT working with the CentOS 6.4 bind package:
[root at server private]# named -V
BIND 9.8.2rc1-RedHat-9.8.2-0.17.rc1.el6_4.4 built with
'--target=x86_64-redhat-linux-gnu' '--program-prefix=' '--prefix=/usr'
'--exec-prefix=/usr' '--bindir=/usr/bin' '--sbindir=/usr/sbin'
'--sysconfdir=/etc' '--datadir=/usr/share' '--includedir=/usr/include'
'--infodir=/usr/share/info' '--with-libtool' '--localstatedir=/var'
'--enable-threads' '--enable-ipv6' '--with-pic' '--disable-static'
'--with-dlz-filesystem=yes' '--with-gssapi=yes' '--disable-isc-spnego'
'CFLAGS= -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions
-fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic' 'CPPFLAGS=
using OpenSSL version: OpenSSL 1.0.0 29 Mar 2010
look at: '--with-gssapi=yes' ' (looks like the compile option is set to
'yes' when its meant to be a directory path) wtf?
shouldn't this be: --with-gssapi=/usr/include/gssapi/'
[root at server private]# ls /usr/include/gssapi
gssapi_ext.h gssapi_generic.h gssapi.h gssapi_krb5.h mechglue.h
[root at server private]#
On Sun, Jun 2, 2013 at 5:40 PM, Gary Maurizi <garymaurizi at gmail.com> wrote:
> I want to thank you both so very much for your help.
> It's another day and I'm back to it, refreshed, and determined to figure
> out what is causing so many issues for the CentOS 6.4 users.
> Going through the same exact steps on ubuntu 12.04 on a different machine
> does give me working dynamic DNS updates, so I have isolated the issue I'm
> having to CentOS only slightly.
> Though bind does not run chrooted by default/at all on CentOS 6.4, I am at
> the point of wondering if maybe some of the samba related features are
> either compiled in and broken/buggy, or not compiled in at all for the bind
> package in the base repositories.
> I would like to try compiling bind 9.9 from source with all of the options
> explicitly stated, but was just wondering if maybe some one could take a
> look at the build options for the CentOS-Base repo version of bind and see
> if anything sticks out as missing, I don't want to miss something samba
> needs in 9.9 using the same options presented below, when I do this. :)
> [root at server samba-master]# named -V
> BIND 9.8.2rc1-RedHat-9.8.2-0.17.rc1.el6_4.4 built with
> '--build=x86_64-redhat-linux-gnu' '--host=x86_64-redhat-linux-gnu'
> '--target=x86_64-redhat-linux-gnu' '--program-prefix=' '--prefix=/usr'
> '--exec-prefix=/usr' '--bindir=/usr/bin' '--sbindir=/usr/sbin'
> '--sysconfdir=/etc' '--datadir=/usr/share' '--includedir=/usr/include'
> '--libdir=/usr/lib64' '--libexecdir=/usr/libexec'
> '--sharedstatedir=/var/lib' '--mandir=/usr/share/man'
> '--infodir=/usr/share/info' '--with-libtool' '--localstatedir=/var'
> '--enable-threads' '--enable-ipv6' '--with-pic' '--disable-static'
> '--disable-openssl-version-check' '--with-dlz-ldap=yes'
> '--with-dlz-postgres=yes' '--with-dlz-mysql=yes'
> '--with-dlz-filesystem=yes' '--with-gssapi=yes' '--disable-isc-spnego'
> '--enable-fixed-rrset' 'build_alias=x86_64-redhat-linux-gnu'
> 'host_alias=x86_64-redhat-linux-gnu' 'target_alias=x86_64-redhat-linux-gnu'
> 'CFLAGS= -O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions
> -fstack-protector --param=ssp-buffer-size=4 -m64 -mtune=generic' 'CPPFLAGS=
> using OpenSSL version: OpenSSL 1.0.0 29 Mar 2010
> using libxml2 version: 2.7.6
> [root at server samba-master]#
> Thank You so much,
> On Sun, Jun 2, 2013 at 4:36 PM, Andrew Bartlett <abartlet at samba.org>wrote:
>> On Mon, 2013-06-03 at 01:11 +0200, steve wrote:
>> > On Mon, 2013-06-03 at 08:16 +1000, Andrew Bartlett wrote:
>> > > On Mon, 2013-06-03 at 00:05 +0200, steve wrote:
>> > > > Hi
>> > > > openSUSE 12.3
>> > > > This is the first time in many years where the SUSE/openSUSE bind
>> > > > _almost_ worked out of the box. They will not entertain non chrooted
>> > > > installs.
>> > >
>> > > This is somehow totally disabled?
>> > No. You can enable it, but the chroot is the default. You cannot install
>> > bind without the bind-chroot environment package too.
>> > >
>> > > > I've tested it. It's OK without tkey-domain nor
>> > >
>> > > Good.
>> > >
>> > > > I am trying to present as minimal a setup for the OP. I think in
>> > > > situations such as these, it is important to get bind working choose
>> > > > what. For that we must cut it down to an absolute minimal install
>> > > > security settings wide open. once it's working, then we can. . .
>> > > >
>> > > > I think that DNS is still our weakest link and I'm really pleased
>> to see
>> > > > the devs looking through the end user list occasionally. Until the
>> > > > internal DNS is ready, we're stuck with bind. Let's try and make it
>> > > > painless as possible for ourselves.
>> > >
>> > > The only way we can really improve it (as far as I'm currently aware)
>> > > to take the bind binary, and launch it with a custom config file
>> > > 'samba' like we do smbd, pointing only at our DNS zone, and with
>> > > etc disabled.
>> > >
>> > > That should, in theory, get us most of the control we get with the
>> > > internal server. Someone needs to write the patches however, and it
>> > > would mean we gain yet another DNS mode (which may be more trouble
>> > > it's worth - I don't know).
>> > >
>> > > Andrew Bartlett
>> > End users need something simple to install. We also need something that
>> > does dynamic dns reliably. The strong points of the internal dns are
>> > it's simplicity of installation. Would it be possible to get it to do
>> > dns updates from nsupdate?
>> It does do dns updates from nsupdate. There is a client-side error
>> shown *after* the successful update, but the developer who developed the
>> patch for this hasn't been able to write the tests to allow his changes
>> to make it into master.
>> > The only reason most of us have to go with
>> > bind is because we need reliable dynamic dns updates. Not just sometimes
>> > and then only with windows clients. Many of the questions and confusion
>> > on this list is to do with DNS. Get that sorted and you have a killer
>> > app.
>> We are not aware that this is anything more than a cosmetic issue. We
>> know it looks really bad, but we need someone to pick up that patch, and
>> find a way to test.
>> > As this is a very big stopper for many of us, would it be possible to
>> > consider a change of developer emphasis for 4.1? Something like a 'DNS
>> > or bust' approach? Many of the things you are doing are amazing but
>> > without the basic DNS, they're lost on us end users. If you wanted any
>> > DNS testers to get it to the rolling out stage, I'm sure many of us here
>> > would be only too pleased to help you test whatever you could throw at
>> > us.
>> Sadly that just isn't how the Samba Team works, sorry.
>> > Thanks for reading. Please don't lose sight of those of us do not code.
>> > We're still very much Samba and still very much here to help the devs
>> > and so the project.
>> We do very much appreciate your interest.
>> Andrew Bartlett
>> Andrew Bartlett
>> Authentication Developer, Samba Team http://samba.org
>> To unsubscribe from this list go to the following URL and read the
>> instructions: https://lists.samba.org/mailman/options/samba
More information about the samba