Re: Mutt displays mail twice... :-s
On Mon, 27 Oct 2003 19:06:20 -0500
"Derrick 'dman' Hudson" <dman@dman13.dyndns.org> wrote:
> On Mon, Oct 27, 2003 at 11:23:35PM +0100, Tom wrote:
>
> | However, the longer this takes, the more I'm beginning to feel a
> | little nervous, since it undoubtedly has to do with some
> | misconfiguration of mine.
>
> Don't be so nervous, you just get duplicate mails. It's not the worst
> that could happen :-).
>
> | Diff for the two messages of the above example spits out this:
> |
> | 2c2
> | < ([127.0.0.1] helo=localhost ident=tom)
> | ---
> | > ([127.0.0.1] helo=localhost ident=fetchmail)
> | 4,5c4,5
> | < id 1AEFQo-0006UD-00
> | < for <tom@localhost>; Mon, 27 Oct 2003 23:01:46 +0100
> | ---
> | > id 1AEFP1-0006St-00
> | > for <tom@localhost>; Mon, 27 Oct 2003 22:59:55 +0100
> | 9c9
> | < for tom@localhost (single-drop); Mon, 27 Oct 2003 23:01:46
> | +0100(CET)
> | ---
> | > for tom@localhost (single-drop); Mon, 27 Oct 2003 22:59:55
> | > +0100(CET)
> |
> | One of the messages seems to be delivered about a minute later than
> | the other, and it's obviously treated as a different message
> | (exim-id's differ). Me and my newbie-ness suspect it has something
> | to do with the first difference (ident=tom <-> ident=fetchmail).
> | Could someone enlighten me...?
>
> Can you post the entire headers for both of those messages? The first
> step will be to trace the Received: headers to see when and where the
> messages travelled. As you've noted, the problem isn't in mutt or
> procmail. The duplication happens before either of those programs
> becomes involved. I suspect you are using fetchmail and for some
> reason it is handing the message to exim twice, thus you get two
> copies. Its also conceivable that the problem lies even earlier than
> that with your mail provider, but we'll find out one step at a time.
>
> Hmm, actually, now that I think about it, I know what the problem is.
> (How nice of exim to log 'ident' information, and how nice of your
> system to provide it! :-D). 'ident', btw, is a mechanism whereby a
> network host can ask another one what user owns the process that has
> the socket open. In this case it really helps identify and solve the
> problem (next paragraph ...).
>
> You have fetchmail running twice -- once as user 'fetchmail' and once
> as user 'tom'. Both instances are grabbing the mail from your POP box
> and passing it on to exim. You have fetchmail set to not remove
> messages from the server, and POP has limited capability of
> identifying "read" messages, so each fetchmail ends up fetching each
> message.
>
> The solution is to clear out your /etc/fetchmailrc. (IMO running
> fetchmail from your user's own crontab is better than running it as a
> system-wide daemon, so I recommend keeping your ~/.fetchmailrc and
> emptying /etc/fetchmailrc.)
>
> -D
That is a common denominator.
All I have is evolution and an unconfigured fetchmail.
Pretty basic because at this stage I'm not capable of more, but I was
double posting for about three days, and then without doing anything
about it, it just cleared up.
Regards,
David.
Reply to: