Hi,

Just received an answer from David:

"Are you using any mailing lists? Each mailing list counts as a single user.
Also, the reserved "postmaster" and "maiser" accounts also count as
users."

Now it's clear.

Cheers

Joerg

> We had mercury crashing about 3 times in the last 6 months but we
> didn't notice until a few days after it hapenned. Is there a way to
> get notifications of it crashing? or is there a external tool we could
> use to let us know it crashed or to make sure that the emails didn't
> get send?

1.    Use something like ServersAlive <http://www.woodstone.nu/salive/> to monitor the server.  It also can send notifications.  There are many others as well but I've not tested them.

2.    I use both the Mercury and NT wrapper to load Mercury as a service, both with automatically restart the server.

aol.com is known to be very strict about the reverse DNS, so this could be the problem. For full details about the connection error you can temporarily switch on session logging in MercuryE.

/Rolf 

No Rolf.  It was me who was missing something!  At the moment, I am saving a message to a file within Global Filtering.  I guess I can have Policy do this instead ... which would probably be simpler (fewer statements in the Global Filtering rules).

Thank you. 

Gordon

why are you want to need a detail? tell me?

The sender address for a message is normally set by the client program that sends the message to Mercury. For some server generated notifications it's created from "Internet name for this system" in Mercury core configuration. (See Mercury help!)

/Rolf 

You will get a separate email with the required licence file. This isn't integrated in the automatic Paypal routine, so there is unfortunately some delay.

/Rolf 

[quote user="PaulW"]

Can any of your other providers handle all your outgoing mail?

[/quote]

Unfortunately no

[quote user="PaulW"]

And why can you not send from Outlook?

[/quote]

We have a strict policy about mail: all incoming and outgoing must be processed in M32, for  archiving, control and filtering purposes.

Anyway, I solved setting up multiple instances of M32.

It would be nice if in future versions of M32 we could use multiple instances of the sole MercuryC (I seem to remember that someone here in the forum said that in Netware it's possible); anyway, I think M32 is a great piece of software; absolutely worth the registration.

With so little information it's impossible to say. Start by verifying that local domains in Core configuration are specified correctly (check Mercury help or the manual). If that isn't it please post the contents of your mercury.ini file.

/Rolf 

Hi,

beside all the problems which are discussed here and which most probably arise due to false configuration of Mercury, I would like to state a successful update from 4.73 to 4.74 on a Windows Server 2003 R2 32bit. Update process proceeded without any errors (as usual).

The one and only conspicious thing which I have experienced is, that Mercury D (POP3 Client) is polling all of our POP3 accounts as fast as never before. It seems David has released all brakes. [;)]

Greetings

Joerg

I think I had this same issue and I was never able to figure out what was wrong but a reinstall did fix it.  

Jim 

PaulW - Four years ago I was playing with XAMPP and at that time it was not intended for use in a production environment.

Since it is almost inevitable that this will happen I would try the following:

put something like this in your rules.mer file

If not header "F" contains "postmaster@targetserver.com" SkipNext ""

If header "T" contains "source@sourceserver.com" SendMessage "source@sourceserver.com:error.txt"

If not header "F" contains "postmaster@targetserver.com" SkipNext ""
If header "T" contains "source@sourceserver.com" Delete ""

you would have to create the file error.txt  with the message you want it to say

An email that was just auto-forwarded to target@targetserver.com was not delivered.  It was probably too big.

I think that will work, perhaps others can expand on my idea.

Thanks!

As this already was answered by Paul Whelan in the Mercury mailing list I'll quote his answer here:

Orphaned *.qdf files are normally associated with a server problem - like

running out of

space. What are their dates?  Do they correspond to a problem?  Can you

confirm that they

weren't delivered?

They are the same format as new mail files (*.cnm) so you could rename them

and deliver

them directly, or put them in you own mailbox and forward them.

When the envelope information in the .QCF file is missing that is probably the best way to handle it. It would be possible to have Mercury core do the delivery by making .101 files (adding envelope information to the beginning of the file and changing the extension to .101) as well. The 101 format is quite simple. It starts with a line containing $$ and the sender address, followed by one or more lines with recipient addresses, and then one blank line before the actual message.

/Rolf 

Well, "OK UID FETCH complete" is obviously not an error, so it would seem that the software that is retrieving messages from MercuryI somehow is out of sync. Try restarting it.

/Rolf 

Ok i found my mistake...i needed to use SSL and not STARTTLS for MercuryC SMTP relay client.

Maybe I found where is the problem.

Here is the QCF/QDF of a test message in the queue folder of the first instance:

[quote]

ST: R 000000000000 00 000001 000000 000000

FR: <pinco.palla@pec.it>

DF: MG000022.QDF

FL: 65536      

OS: 120112111502

BA: <myhouse@blabla.it>

ES: R 120112111502

RI: 000.000.000.000 000.000.000.000 000.000.000.000 000.000.000.000

DI: --------.---

EA:

Return-path: <pinco.palla@pec.it>

Received: from ugenerico2 (192.168.1.8) by palla09.local (Mercury/32 v4.72) with ESMTP ID MG000022;

   12 Jan 2012 11:15:02 +0100

From: "Pinco Palla" <pinco.palla@pec.it>

To: <myhouse@blabla.it>

Subject: prova 12genn h.11.14

Date: Thu, 12 Jan 2012 11:15:02 +0100

Message-ID: <00f201ccd113$0c3d9cd0$24b8d670$@palla@pec.it>

MIME-Version: 1.0

Content-Type: multipart/alternative;

    boundary="----=_NextPart_000_00F3_01CCD11B.6E0204D0"

X-Mailer: Microsoft Office Outlook 12.0

Thread-Index: AczREwwXeQjdos+cQ1+Jps9WknKHzw==

Content-Language: it

...cut......cut......cut...

[/quote]

As already seen, this message is moved by an outgoing rule to the

PEC-out mailbox, and sits there waiting to be picked up by Wsmtpex; the

following is the Header of the message found in the CNM file in the

PEC-out folder

[quote]

Received: from spooler by palla09.local (Mercury/32 v4.72); 12 Jan 2012 11:17:50 +0100

Received: from ugenerico2 (192.168.1.8) by palla09.local (Mercury/32 v4.72) with ESMTP ID MG000022;

   12 Jan 2012 11:15:02 +0100

From: "Pinco Palla" <pinco.palla@pec.it>

To: <myhouse@blabla.it>

Subject: prova 12genn h.11.14

Date: Thu, 12 Jan 2012 11:15:02 +0100

Message-ID: <00f201ccd113$0c3d9cd0$24b8d670$@palla@pec.it>

MIME-Version: 1.0

Content-Type: multipart/alternative;

    boundary="----=_NextPart_000_00F3_01CCD11B.6E0204D0"

X-Mailer: Microsoft Office Outlook 12.0

Thread-Index: AczREwwXeQjdos+cQ1+Jps9WknKHzw==

Content-Language: it

...cut......cut......cut...

[/quote]

We can see that the "Return-path: <pinco.palla@pec.it>" line disappeared.

Now, as stated in the wsmtpex.txt file

[quote]

WSMTPEx looks into .CNM file and search for header 'Return-Path' for

SMTP envelope MAIL FROM address. Uses "<>" if none is found.

[/quote]

So, I was thinking that this may lead to the QCF with the empty "FR: <>" field in the queue of the second instance (and to the subsequet problems).

May this be so?

I'm wondering what to do... Any idea?

Best regards

Filippo