Bind 9.2.3 very, very slow starup

[Eric Dynowski]

Everyone,

Just an update..

I was about to rebuild bind on another machine and replace this one
because I was ran out other ideas. So I tar'ed up the /etc/bind direcory
to copy over, and noticed the copy (scp) was very, very, very slow.. So I
disabled the on board Intel NIC, installed a Netgear, and recompiled the
Kernel (2.4.18) without EtherExpress Pro support and with DEC Tulip
support, then rebooted and bind started perfectly the first time.

Because this is a production server, Im still affraid to restart bind and
find out if this really solved the problem. I guess we will see what
happens the next time I try and restart. Has anyone ever seen a problem
with bind relating to network kernel drivers? I have tried different
kernel versions up to 2.6..

Strange..

Eric.

> On Tue, 2004-07-13 at 19:20, Eric Dynowski wrote:
>> > At 01:46 PM 7/13/2004, Eric Dynowski wrote:
>> >>Danny,
>> >>
>> >>There are no errors in the logs, so I am assuming there are none in
>> the
>> >>zone files. I have removed all zones but one and still the same
>> problem
>> >>occurs.
>> >>
>> >>Eric.
>> >
>> > run named-checkzone to make sure and named-checkconf.
>> >
>> > Danny
>> >
>> >
>>
>> Danny,
>>
>> named-checkzone, and named-checkconf return all OK. No problems found.
>>
>> I ran named-checkzone on all of our zone files.
>>
>> Eric.
>
> One thing comes to mind, are you starting named *BEFORE* you have your
> network connections up and active?
>
> --
> G. Roderick Singleton <gerry at pathtech.org>
> PATH tech
>
>