[exim] Exim, Vacation and Horde

Αρχική Σελίδα
Delete this message
Reply to this message
Συντάκτης: Mike Jones
Ημερομηνία:  
Προς: 'exim-users@exim.org'
Αντικείμενο: [exim] Exim, Vacation and Horde
Hi,
I've been trawling the internet for the past couple of days to try and find an answer to a little problem I have, but none of the solutions seem to fit what I'm trying to do.

Basically I have Exim, Horde and Imp installed with version 1.1.9 of vacation sitting in /usr/local/bin on a FC 6 server.

Everything works fine, except for the vacation module of Horde which generates a .forward file of

\user, "|/usr/local/bin/vacation user"

Apart from the fact that the .forward file is missing the line # Exim filter, Exim doesn't seem to like that command (in fact no reply is sent, and emails don't get delivered).

However with a bit of routing around, I found that this command does work in the .forward file if put in manually.

unseen pipe "/usr/local/bin/vacation -a user-alias user"

So, my question is simply this - why does the command generated by horde (and by vacation at the command line) not work with exim, but the command I discovered does work ?

The configure file is right out of the box (with a few tweaks for courier-imap) with no routers etc for vacations, so my understanding is that the .forward file is being handled by the forward router and transport.

Any help gratefully received

MikeFrom temmokan@??? Thu May 08 10:09:38 2008
Envelope-to: exim-users@???
Received: from wa-out-1112.google.com ([209.85.146.182]:32888)
    by tahini.csx.cam.ac.uk with esmtp (Exim 4.69)
    (envelope-from <temmokan@???>) id 1Ju285-0000a9-PP
    for exim-users@???; Thu, 08 May 2008 10:09:37 +0100
Received: by wa-out-1112.google.com with SMTP id m28so831970wag.8
    for <exim-users@???>; Thu, 08 May 2008 02:09:32 -0700 (PDT)
Received: by 10.114.13.1 with SMTP id 1mr2860242wam.4.1210237772420;
    Thu, 08 May 2008 02:09:32 -0700 (PDT)
Received: by 10.115.108.15 with HTTP; Thu, 8 May 2008 02:09:32 -0700 (PDT)
Message-ID: <70600bcb0805080209i3f43b35cg96cb577da4d0022c@???>
Date: Thu, 8 May 2008 16:09:32 +0700
From: "Konstantin Boyandin" <temmokan@???>
To: exim-users@???
MIME-Version: 1.0
Content-Type: text/plain; charset?O-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline
X-Spam-Score: -1.8 (-)
X-Spam-Status: No, score?.8 required~0 tests?L?.332,
    BAYES_00?.5 autolearnO version^1.8
Subject: [exim] Exim questions: allow mail from unroutable addresses,
    pipe transport question
X-BeenThere: exim-users@???
X-Mailman-Version: 2.1.9
Precedence: list
List-Id: A user list for the exim MTA <exim-users.exim.org>
List-Unsubscribe: <http://lists.exim.org/mailman/listinfo/exim-users>,
    <mailto:exim-users-request@exim.org?subject?subscribe>
List-Archive: <http://lists.exim.org/lurker/list/exim-users.html>
List-Post: <mailto:exim-users@exim.org>
List-Help: <mailto:exim-users-request@exim.org?subject?lp>
List-Subscribe: <http://lists.exim.org/mailman/listinfo/exim-users>,
    <mailto:exim-users-request@exim.org?subject?bscribe>
X-List-Received-Date: Thu, 08 May 2008 09:09:38 -0000


Hello,

First, if anyone is interested, I managed to solve the problem of
using DSPAM when it is vital that all the possible messages duplicates
be eliminated *before* the messages are processed by spam-filter and
delivered.

It was done by using transport_filter, of course. It's not much of a
walkaround (since I don't think that making spam-filtering via a
router is that more correct).

1. However I would like to ask those knowing the Exim intrinsics: if
a router is used as intermediate message processing (like it was in my
case), is it inevitable that message duplicates will occur?

I explain the problem: we have several 'multi-forward' mail addresses
of the kind:

multibox@???: multibox,<list of other recipients>

A message can be sent to many such multi-forwarders simultaneously (in
To/Cc fields) and it is essential that in such cases a person receive
a single copy of the same message.

However, if pipe transports beget a new instance of Exim they call to
continue message processing, then duplicates become inevitable - I
couldn't exclude them until I eliminated the router/pipe transport
pair to preprocess messages.

2. Also, if someone could provide an advice on how to allow just
several unroutable (unverifiable) sender addresses, I would be glad.

The problem: we receive automated messages with fake (non-existing)
sender addresses. We have no control over the sender's network and I
wouldn't like to rewrite the addresses just to make them real ones -
they must be left inatct. But I won't to allow unroutable addresses
from anywhere as well.

is it possible to allow just selected unroutable addresses and deny
all the rest?

Thanks.