Understood - it's definately an odd issue.

On a separate dedicated Windows 2000 AdvSrvr/IIS 5.0 relay machine what I
was seeing was every few days the IIS service would lock up and mail would
not be transferred. On that machine I tried un-installing IIS and
re-installing it but I had the same issue. This seems more in keeping with
Lance's original post.

Rather then try and isolate the issue on that particular machine I just
changed my email structure around and went back to using another server,
this one a dedicated Windows 2000 AdvSrvr machine that only runs Exchange
2000 EE.

What I'm seeing is something that I suspect is a variant of the same issue.
With 5xxsink bound to a second VS (strictly for sending mail) I get errors
every few days as follows;

Event ID:7031
The IIS Admin Service service terminated unexpectedly. It has done this 1
time(s). The following corrective action will be taken in 1 milliseconds:
Run the configured recovery program.

Event ID: 7031
The Microsoft Exchange IMAP4 service terminated unexpectedly. It has done
this 1 time(s). The following corrective action will be taken in 0
milliseconds: No action.

Event ID: 7031
The Network News Transport Protocol (NNTP) service terminated unexpectedly.
It has done this 1 time(s). The following corrective action will be taken in
0 milliseconds: No action.

Event ID: 7031
The Microsoft Exchange POP3 service terminated unexpectedly. It has done
this 1 time(s). The following corrective action will be taken in 0
milliseconds: No action.

Event ID: 7031
The Microsoft Exchange Routing Engine service terminated unexpectedly. It
has done this 1 time(s). The following corrective action will be taken in 0
milliseconds: No action.

Event ID: 7031
The Simple Mail Transport Protocol (SMTP) service terminated unexpectedly.
It has done this 1 time(s). The following corrective action will be taken in
0 milliseconds: No action.

Event ID: 7031
The World Wide Web Publishing Service service terminated unexpectedly. It
has done this 1 time(s). The following corrective action will be taken in 0
milliseconds: No action.

The good news is that unlike the dedicated IIS 5 machine the Exchange server
quickly re-starts those services (it looks like within about 10 seconds
altogether) and mail works again for several days.

Unbinding 5xxsink causes the issue to stop appearing. What I've also found
is that occassionally spam email will actually make it through 5xxsink -
even though the destination email address does not exist and is not in
either the email address list or the domain list.

However, 99.99% of every other attempt to send mail to the same non-existant
email address is 550'd and the connection is dropped. So for some reason,
and I can't see why (yet), the odd time an email makes it through even
though that address get's 550'd every other time.

The time period that this occurs does not co-incide with when those services
are re-starting. In fact, I can't see any pattern at all when that odd piece
of mail gets through. In my case the mail server is radically
under-utilized.

In my travels I did come across this
(http://www.eventid.net/display.asp?eventid=7031&eventno=465&source=Service%
20Control%20Manager&phase=1);

Yiannis Papadopoulos (Last update 6/30/2003):
- Service: IISAdmin, IMAP4, NNTP, POP3, MSE Routing Engine, SMTP, WWW.
Appeared after installing an SMTP event sink for MS Exchange 2000.
Resolution: unregister the sink. If this fails unregister the DLL of the
sink and restart Windows. See M313404 on how to register an SMTP event sink.

and it got me thinking that the only constant variable (that I can think of)
is 5xxsink was installed on both servers that exhibited this issue.

In my case, I think I'm probably going to leave it alone on the Exchange
machine due to the fact that it quickly re-starts the services - and 5xxsink
is highly valued - even if there's a glitch somewhere the benefits far
outweigh the issue.

In hindsight though, I probably could have configured those services to
automatically re-start on the original relay machine and kept it in service
but I thought there might be something up with it.

I'll keep looking into it to see if there's something else I can come up
with but I suspect the issue lies deeper then my abilities to resolve it at
this point.


[quoted text, click to view]

[quoted text, click to view]

Yeah, guys, I wish I could go further with you, but I don't know what
combo of settings could be causing this without talking a deep (and
destructive) look at your servers. I can't replicate this far-reaching
crash. Sorry.

--Sandy


------------------------------------
Sanford Whiteman, Chief Technologist
Broadleaf Systems, a division of
Cypress Integrated Systems, Inc.

asdfgafhg
[quoted text, click to view]


begin 666 hi guys.txt
)87-D9F=A9FAG
`
end