I have 3 machines and all have been updated to 10.2.8 and all work fine.
I did notice that this problem occurred after the last security update.
In my case that was a temporary action as everything came back to
normal after a few minutes.
Here is the topography of my network.
I am running a MacOSX Server 10.2.8. I have a DSL line with a DSL modem.
I have a LinkSys behind of this DSL modem that is taking care of my
intranet.
My intranet is in the class C network, therefore 192.168.1.0 range to
192.168.1.255.
I do not really host DNS Bind. I just have it set to resolve my
intranet and one WAN domain.
Pacbell, my ISP provider, host the real DNS and is providing the
reverse for my WAN domain.
Just as a remind, for you to be able to host Bind yourself, you must
have 2 different networks where you can setup your NS1 and NS2 TTL
files. Even thou I have 2 machines and I do have one setup as NS1, and
the other set up as NS2, still that doesn't qualify me to host a real
DNS setup, if you got the point here.
The only thing I am doing is to take care of my intranet thru the 4
port LinkSys router.
Sure I am also managing my own firewall.
If that apply to you, let see if we can share the info in our network
and try to resolve this problem.
Here goes a suggestion that you also going to find it in other lists.
Do not be afraid to give as much information about your problem by even
posting your files on line.
That is the same advise all the BIND list gurus will give to you.
Believe me, it will save time and you will
going to get your problem resolved faster. Hiding vital information
just make things harder to resolve.
Eduardo
PS: If your info is taking to much memory and we all know that this
list have a memory restriction in posting large files, send it directed
to those involved in resolving this problem. I just think that, if in
the end we end up to resolve the issue, we must share the results with
the list so others can learn from it. Thanks!
On Sunday, Oct 31, 2004, at 09:46 US/Pacific, David Thompson wrote:
I have the same problem. I have the same server setup identically and
have never changed it. It seems to me that it's from the 10.3.5 update
as this did not occur under the 10.3.4 server version. It is no a
problem with just Safari, as both I.E. and FireFox give the same page
not found error as that of Safari.
I have even gone as far as to backup the DNS zones I have and rebuild
from scratch. Since they are small zones they are easily re-creatable.
Upon hitting the enter key in a browser again after the first page not
found error, the site loads very quickly. The very weird thing is that
if I try this on a windows box, it runs fine and always finds the page
being looked for.
I have done the following with no changes to the system:
-lookupd -flushcache (both on server and on clients)
-changed the named.conf files to no longer be able to dynamically
update the zones
-downloaded the latest root server files (named.ca)
-Deleted the localhost zones files
-Deleted the main zones files
-Rebuilt the zone files, rebooted server
-Using ISP's DNS servers allows for quick loading and resolution of
sites being browsed to
Seems to me it's a problem with the DNS server itself. BIND is version
9.2.2 (which I would assume is the same as everyone elses here)
~
David Thompson
Apple Consultants Network
416-833-3944
ACTC 10.1 ACTC 10.2 ACTC 10.3
On 30-Oct-04, at 11:30 PM, Eduardo Bonsi wrote:
--
Eduardo Bonsi
IT Network Admin
Web Developer
email@hidden
www.beart.com
