Hi All,

I'm having a horrible time only with mass mailouts (with
recipients in the BCC field), and get a ton of "5.7.1
unable to relay for" type NDRs.  Normal day-to-day email
works just fine.

When looking at the event logs, I'm finding that the SMTP
service is actually trying to deliver some of these
emails to webservers and other A record type hosts,
instead of MX hosts!  Wa wa what???

Then, I noticed a post in Google from Microsoft saying
that "If an MX record is not found Exchange will attempt
to deliver to an A record. This tells me the Exchange
server is querying DNS an not getting the correct
answer." (see below)

(http://groups.google.com/groups?
hl=en&lr=&safe=off&threadm=bX8DT%230aEHA.2804%
40cpmsftngxa06.phx.gbl&rnum=1&prev=/groups%3Fq%3Dmx%
 2Bunable%2Brelay%2Bsmtp%2Bwindows%2B2003
%26hl%3Den%26lr%
 3D%26safe%3Doff%26selm%3DbX8DT%25230aEHA
.2804%
2540cpmsftngxa06.phx.gbl%26rnum%3D1)

Because the mailout is so big, it would seem to me like
the SMTP service isn't giving itself enough time to grab
these MX records, and opting for the A records instead.
When I do an NSLOOKUP from the server itself, it's able
to grab the MX record without any problems... plus, I can
telnet over port 25 without any problems to these MX
hosts...

Does anyone know if there is there any way to force the
SMTP service to only look for MX records, and not fall
back to A records... or perhaps increase the timeout
before falling back to A records?

To explain my environment... I have a caching-only DNS
server (not using forwarders) running on this Win2k3/IIS6
machine to speed up DNS queries... It's also sitting
behind an ISA 2000 server firewall.  The ISA server is
hooked up to a T1.

I appreciate your response immensely!

[ Post a follow-up to this message ]

> To explain my environment... I have a caching-only DNS
> server (not using forwarders) running on this Win2k3/IIS6
> machine to speed up DNS queries... It's also sitting
> behind an ISA 2000 server firewall.  The ISA server is
> hooked up to a T1.


What does your DNS cache if it does not forward?

[ Post a follow-up to this message ]

It just does lookups... and caches those lookups.  I'm
not forwarding to ISP DNS servers out there...

>-----Original Message-----
> 
Win2k3/IIS6[vbcol=seagreen] 
>
>
>What does your DNS cache if it does not forward?
>
>
>.
>

[ Post a follow-up to this message ]

But you should do so....

<anonymous@discussions.microsoft.com> wrote in message
news:0a6901c4abaa$3ddcff60$a401280a@phx.gbl...[vbcol=seagreen]
> It just does lookups... and caches those lookups.  I'm
> not forwarding to ISP DNS servers out there...
> 
> Win2k3/IIS6 

[ Post a follow-up to this message ]

But why?  Windows DNS has the capability of using root
DNS servers... why not use them?  Is Windows in that bad
of a shape that you have to use some ISP's unix DNS boxes
to do standard DNS lookups?

>-----Original Message-----
>But you should do so....
>
><anonymous@discussions.microsoft.com> wrote in message
>news:0a6901c4abaa$3ddcff60$a401280a@phx.gbl... 
DNS[vbcol=seagreen] 
is[vbcol=seagreen] 
>
>
>.
>

[ Post a follow-up to this message ]

On Wed, 6 Oct 2004 08:55:09 -0700, "vnvjeep" <vnvjeep@hotmail.com>
wrote:

>But why?  Windows DNS has the capability of using root
>DNS servers... why not use them?  Is Windows in that bad
>of a shape that you have to use some ISP's unix DNS boxes
>to do standard DNS lookups?

No, you don't need to have forwarders if you have access to the root
servers.

Jeff
[vbcol=seagreen] 
>DNS 
>is 

[ Post a follow-up to this message ]

Performance wise requests should not sent directly to the rootservers.
And as you see in your queue - you have network performance problems.....

Not Windows is to blame for that but your configuration

"vnvjeep" <vnvjeep@hotmail.com> wrote in message
news:00b301c4abbc$db6f6ad0$a301280a@phx.gbl...[vbcol=seagreen]
> But why?  Windows DNS has the capability of using root
> DNS servers... why not use them?  Is Windows in that bad
> of a shape that you have to use some ISP's unix DNS boxes
> to do standard DNS lookups?
> 
> DNS 
> is 

[ Post a follow-up to this message ]