slow IXFR updates

Chris Buxton cbuxton at menandmice.com
Mon Aug 21 17:08:31 UTC 2006


Sounds to me like a problem with DNS notify. Is the customer  
correctly referencing your name server in an NS record in each of his  
zones?

As a quick and dirty workaround, have the customer add the following  
to his options statement:

	also-notify { ip-addr; };

where "ip-addr" is the IP address of your name server. If he already  
has an also-notify substatement, have him instead add your name  
server to the existing list. Note that this can be overridden in each  
zone statement, so if he has also-notify substatements in his zone  
statements, he'll need to modify those instead.

If that solves the problem, then you know it was due to a problem  
with DNS notify. Then you and your customer can set about fixing that  
the correct way (making sure the NS records and SOA record mname are  
correct), because the problems that cause DNS notify to fail often  
cause other problems as well.

Chris Buxton
Men & Mice

On Aug 19, 2006, at 9:21 AM, Rainer Duffner wrote:

> Hi,
>
> I run a FreeBSD6.0 BIND 9.3.1 in a jail as a secondary-only for  
> various
> customers.
> Now, one of these customers complains that the updates take very long
> (1-3 hours).
>
> What could be that reason for that?
> What would I need to change in the configuration?
> The only thing I did to the configuration is to disable recursion and
> remove or comment out the various examples.
> The host is running without a firewall currently, so that should  
> not be
> a problem.
>
> Any ideas?
>
>
> Best regards,
> Rainer
>
>
>



More information about the bind-users mailing list