[CentOS] problem with slave dns servers

Fri Aug 29 04:26:49 UTC 2008
Mark Quitoriano <markquitoriano at gmail.com>

On Fri, Aug 29, 2008 at 6:27 AM, Robert Spangler
<mlists at zoominternet.net> wrote:
> On Thursday 28 August 2008 05:50, Mark Quitoriano wrote:
>
>>  hmmm... yeah i think is et everything to 300 which is not good. What
>>  is the recommended TTL settings? some sites recommend 4 days some 1
>>  hour.
>>
>>
>>  On Thu, Aug 28, 2008 at 3:17 PM, Michel van Deventer
>>
>>  <michel at van.deventer.cx> wrote:
>>  > Hi,
>>  >
>>  > what is the TTL of your DNS records ? After TTL expires, the slaves
>>  > don't respond to queries either, because the records aren't valid
>>  > anymore. If your primary is down longer than the TTL of your DNS records
>>  > you could reconfigure one of the slaves as a new primary or maybe
>>  > consider making more than 1 primary.
>>  >
>>  > On Thu, 2008-08-28 at 09:01 +0200, Romeo Ninov wrote:
>>  > > Have you describe all the slave servers in you domain configuration
>>  > > (in registrant)?
>>  > >
>>  > > Mark Quitoriano  wrote / napísal(a):
>>  > > > Hi,
>>  > > >
>>  > > > I have 4 bind9 dns installed on centos 4. My primary dns server went
>>  > > > down and all of my domains doesn't resolve even if the 3 slave dns
>>  > > > is up and running. Im not sure where to configure this is it in my
>>  > > > domain registration or in bind?
>
> It is not your TTL values that is the problem it is the EXPIRE value.
>
> TTL is used for the caching of the information and tells the cache when to
> remove the information.
>
> EXPIRE is the value that tell the slave how long the information it the zone
> file is good for when the master cannot be reached.
>
> Leave your TTL at 300 (5 min) and change the EXPIRE value to something like a
> week or more.
>
>

Ok that clarifies everything :) thanks man!