Re: [Exim] message_size_limit and error-messages

Top Page
Delete this message
Reply to this message
Author: Jeffrey Goldberg
Date:  
To: Philip Hazel
CC: Sven Callender, Exim Users
Subject: Re: [Exim] message_size_limit and error-messages
On Wed, 22 Sep 1999, Philip Hazel wrote:

> On Wed, 22 Sep 1999, Sven Callender wrote:
>
> > Now the Management has had a new idea: Exim shall send a Message to
> > the original recipent telling him/her that a message was "blocked" due
> > to the size.
>
> That shouldn't be necessary. If Exim has refused a message because of
> the size, whatever was trying to send the message to Exim should be the
> thing that constructs the error message and sends it back.


I think that the question was to inform the recipients, not the sender,
who will get an appropriately generated bounce. It appears that the
management want mail that says something like

"Someone tried to send you mail, but it didn't get through because the
sender probably thinks that MS-Word or PowerPoint are reasonable
document exchange formats. Now do you really want to be getting mail
from such idiots anyway?"

or something like that, if I've understood the original message.

Anyway, what they want is hard because most size rejections should happen
very early if I understand correctly, because the client could respond
intellegently to the SIZE reported by the initial EHLO. In that case,
there will be no information that exim has to do anything with.

Other size rejections (which really should be the minority if I
understand) will happen late so that at least exim has the information.
The information needed (recipients) is, I believe, logged in the
exim_rejectlog. Still I doubt that there is a way, even for these, to do
what you want (or what your bosses want) directly in exim. Best bet would
be to run a script that reads the reject log and generates mail from that.
But that seems pointless in the light of the fact that you'll only get a
small portion of size rejected messages. Or am I mistaken about how the
SIZE stuff works? (Sorry, I haven't actually read the relevant RFCs, I'm
just guessing from what I think make sense.)

-j

-- 
Jeffrey Goldberg                +44 (0)1234 750 111 x 2826
 Cranfield Computer Centre      FAX         751 814
 J.Goldberg@???     http://WWW.Cranfield.ac.uk/public/cc/cc047/
Relativism is the triumph of authority over truth, convention over justice.