Outgoing messages are stuck in queue. I belive it's a DNS problem.

[Koichi Senada]

Hi all! I have a problem, outgoing messages are stuck in queue and are not
being delivered to the recipients.

I have tested that Exchange 2000 Server with SmtpDiag, and here's what it
shows:



--------------------------------------------------------------------------------
Searching for Exchange external DNS settings.
Computer name is SRV0.
VSI 1 has the following external DNS servers:
195.222.130.67,195.222.142.34
Checking SOA for perm.ru.
Checking external DNS servers.
Checking TCP/UDP SOA serial number using DNS server [195.222.130.67].
TCP test succeeded.
UDP test failed.
Serial number: 2006100402
Checking TCP/UDP SOA serial number using DNS server [195.222.142.34].
TCP test succeeded.
UDP test failed.
Serial number: 2006100402

Checking internal DNS servers.
Checking TCP/UDP SOA serial number using DNS server [10.0.1.1].
Failed: DNS server [10.0.1.1] may be down.
Checking TCP/UDP SOA serial number using DNS server [127.0.0.1].
Failed: DNS server [127.0.0.1] may be down.
Checking TCP/UDP SOA serial number using DNS server [195.222.130.67].
TCP test succeeded.
UDP test failed.
Serial number: 2006100402
SOA serial number match: Passed.
...


--------------------------------------------------------------------------------

And then MX records were checked, and the program has connected to the
remote port 25.

195.222.130.67 is our ISP's IP-address of their DNS-server, which we are
using. I'm wondering why our own DNS-Server doesn't answer to the
DNS-queries.

Well then I've opened TELNET and sent a testing email to an outside server,
it was send and delivered correctly at the moment.

So I see, that our server connects to outside servers on port 25 normally
through telnet, but the Exchange 2000 Server doesn't do that.

What can I do to figure out the problem?

I am not so smart about DNS, and that's the only part of it which wasn't
checked by me properly. Can anybody help me?

[Koichi Senada]

I have just installed ExTRA and here's what it reports:

---------
Area: Exchange Gateway
Server: srv0
SMTP instance with no outbound SMTP connector set
Server: srv0
It appears that no SMTP connectors with an address space have SMTP instance
Default SMTP Virtual Server as its designated local bridgehead server.
Outbound messages may need another hop before going out of the Exchange
organization. Please check if this configuration is intentional.
---------

I just don't get it, what's to be configured yet, since everything was
working a few weeks ago, and I wasn't changing anything.

[Koichi Senada]

I am glad to report that I have solved the problem.

The DNS server was configured to accept queries only from predefined IP
addresses of the server itself (DNS server, Exchange 2000 Server, Domain
Controller and everything else on one machine).

When some VPN client has connected to the server, a new additional IP which
wasn't listed in DNS server was taken, and since then all queries from
Exchange, and from nslookup, and from every other application were made as
from that new IP. That's why they were declined, and that's why Exchange
couldn't send out any message.



Right now I am checking out "Routing and Remote Access" to find out how to
get rid of such a strange behavior when a machine takes an additional IP.
Probably command prompt applications NETSH and ROUTE will help me with that.



I hope that this report will be userful for the readers. :)