I haven't seen it for years but in the old days
when SMTP was "new" the duplication would sometimes
happen if the sequence of closing the connection
was interrupted.

    There was even an RFC on this! RFC 1047, "DUPLICATE
MESSAGES AND SMTP" from 1988. Of course your problem could
differ but if you are curious see
    http://tools.ietf.org/rfc/rfc1047.txt

    I'm not sure why this would show up on messages
with external recipients (or blocks but each one is a
message with up to 100 RCPT TOs). However, lots has
changed since 1988 - there were usually no spam/virus
scanners and such - so maybe something else is creating
a similar effect. There was also one firewall that did
funky things that caused duplicated e-mail as I recall.
But that too was a long long time ago.

    Of course this may not be the cause of the problem
here but it was fun to reminisce about the old days. ;-)

    marty



Nathan Brindle wrote:
> At 04:51 PM 4/2/2009 -0400, you wrote:
>> If that was true, why would it retry the message with only one of the 8
>> original blocks?
>
> Because the spool files hold one block of 100 addresses each.  It's only one spool file that's involved in each case.
>
>> Why would those duplicate blocks always involve outside addresses?
>
> Couldn't say, but in all fairness that sounds like an issue on sendmail's side, if indeed it's even germane to the problem.
>
>> Would not be the retry recorded in listserv log?
>
> No, because if you have SMTP_FORWARD_n configured, all the outbound mail from LISTSERV is handled by asynchronous "worker" processes, which -- under unix -- don't keep logs.  When you see the message in the LISTSERV log that says the mail has been sent, in reality it has been written to the spool and handed off to the workers for delivery.
>
>> If not, where would it be logged?
>
> The only place you can really find any information about this is going to be in the sendmail log.
>
> I'm partial to this theory myself:
>
>> It sounds like the sendmail server is not confirming the receipt of the
>> message with a 250 Message Accepted, so Listserv is not sure it went
>> through and retries the message.
>
> I agree.  I can't imagine why LISTSERV (or, more properly, the asynchronous SMTP "worker") would ever retry a spool file if the 250 had arrived and the transaction had been properly closed out.
>
> Nathan
>