fetchmail-friends
[Top] [All Lists]

[fetchmail]wish? bug? another case where fetchmail just stops working

2004-09-12 13:14:45
This is similar to bug #223933 & #235379 -- similar in that the same effect 
occurs [fetchmail basically stops retrieving messages] but the cause this 
time appears to be a mal-formed "from" address [and so far, 100% of these 
have been spam]

fetchmail's log reports:
============================================================
fetchmail: 56 messages for <deleted> at mail2.1stnetusa.com (209077 octets).
fetchmail: reading message <deleted>@mail2.1stnetusa.com:1 of 56 (5946 octets) 
fetchmail: SMTP error: 501 Bad address syntax
fetchmail:  not flushed
fetchmail: client/server protocol error while fetching from  
mail2.1stnetusa.com
fetchmail: Query status=4 (PROTOCOL)
fetchmail: sleeping at Sun Sep 12 12:00:30 2004
============================================================

logging in via telnet to "clear" this problem, I see:

============================================================
osnut:/var/log # telnet mail2.1stnetusa.com pop3
Trying 206.171.253.49...
Connected to mail2.1stnetusa.com.
Escape character is '^]'.
+OK POP3 mail2.1stnetusa.com v2003.83rh server ready
user <deleted>
+OK User name accepted, password please
pass <deleted>
+OK Mailbox open, 56 messages
retr 1
+OK 5946 octets
Return-Path: <P[5-](_at_)epomail(_dot_)com>
Received: from cpe-68-116-200-25.ma.charter.com 
(cpe-68-116-200-25.ma.charter.com [68.116.200.25])
        by mail2.1stnetusa.COM (8.12.11/8.12.10) with SMTP id i8C5AVnr031168;
        Sat, 11 Sep 2004 22:10:36 -0700
X-Message-Info: SV/t+480+d/VBT+6/13411299508938
Received: from smtp-cress(_dot_)botulism(_dot_)P[5-](_at_)epomail(_dot_)com 
([68.116.200.25]) by 
q610-pf1(_dot_)P[5-](_at_)epomail(_dot_)com with Microsoft 
SMTPSVC(5.0.3490.1290);
         Sun, 12 Sep 2004 09:50:14 +0400
Received: from signal151(_dot_)aphasia(_dot_)P[5-](_at_)epomail(_dot_)com 
(genevieve158(_dot_)P[5-](_at_)epomail(_dot_)com [68.116.200.25])
        by smtp-uncle(_dot_)spouse(_dot_)P[5-](_at_)epomail(_dot_)com (Postfix) 
with SMTP id 
426DLQ98F19VK
        for <denis(_at_)1stnetusa(_dot_)com>; Sun, 12 Sep 2004 10:45:14 +0500
Received: from smtp-carborundum(_dot_)coupon(_dot_)P[5-](_at_)epomail(_dot_)com 
([68.116.200.25]) by 
zf1-l24(_dot_)P[5-](_at_)epomail(_dot_)com with Microsoft 
SMTPSVC(5.0.6688.9727);
         Sun, 12 Sep 2004 00:50:14 -0500
X-Message-Info: OBSPP+%ND_LC_CHAR[1-3]97+zzb+N+934/953053631257
Received: (qmail 15008 invoked by uid 47); Sun, 12 Sep 2004 04:44:14 -0100
Date: Sun, 12 Sep 2004 10:43:14 +0500
Message-Id: <550025679(_dot_)95613(_at_)P[5-]@epomail.com>
From: Rachel  Melendez <P[5-](_at_)epomail(_dot_)com>
To: Denis <denis(_at_)1stnetusa(_dot_)com>
Subject:
MIME-Version: 1.0 (produced by ranklegauguin 1.8)
Content-Type: multipart/alternative;
        boundary="--0297487016671684712"
Status: RO

----0297487016671684712
Content-Type: text/html;
        charset="iso-9805-0"
Content-Transfer-Encoding: quoted-printable
Content-Description: transfer airedale tolerant

<html>
[rest of message deleted] 
============================================================

Note the FROM line:

     From: Rachel  Melendez <P[5-](_at_)epomail(_dot_)com>

given that this is probably illegal syntax, I gather it is simply a mechanism 
the spammer is using to avoid nastygrams in return.  The net effect, of 
course, is that fetchmail simply "stops" retrieving messages [from an 
end-user's viewpoint -- technically, fetchmail continues to wake up every 5 
minutes, but refuses to budge on this message]

WHAT I WOULD LIKE: a configuration option to drop these messages (delete from 
server) since there isn't really anything that can be done about them.  In 
the case of the original bugs noted, one was due to a hard failure of the 
server to deliver the message, which may mean the message is already gone; 
and the other has to do with a similarly-munged message-ID, which causes a 
headache for fetchmail itself it seems [I would venture to guess the 
duplicate-search is looping, but that's only a guess...]

I'm running fetchmail as a daemon in a basic straight-fetch-and-deliver mode; 
version 6.2.3 from a stock SuSE 9.0 install (and the NEWS file in the latest 
source does not give any indication that this problem has been addressed)

Attachment: pgpJNkFqVOaMs.pgp
Description: signature

<Prev in Thread] Current Thread [Next in Thread>