[milters] Archive

Lists Index Date Thread Search

Article: 290
From: Anthony Howe
Date: 2004-12-23 09:05:59 -0500
Subject: Re: milter-sender not bouncing messages

Removal...........: milters-request@milter.info?subject=remove
More information..: http://www.milter.info/#Support
--------------------------------------------------------

Jeff Powell wrote:
> I'm currently using milter-sender .62 with sendmail 8.13.1. It had been working
correctly but now I notice that it is not bouncing messages that fail the callback.  Here
is a log sample:

This log output doesn't help so much. You need -v info,trace,dialog at 
teh very least so that the function enter/exist points can be seen and 
other info.

I think this is an interaction issue with when and how sendmail handles 
  the milters and rulesets; are you using FEATURE(`delay_checks')?

If the black list hadn't pulled out a result, the milter would have 
rejected based on the dialogue result. The trace information would have 
shown me the exit code from the mxCallback() and filterMail().

Hmmm. I bet you have DeferMailReject=1, then its very possible to see 
the callback result defered until filterRcpt() as intended, but to then 
have a blacklist reject the connection before the milter does.

Actually this is most likely. Here is a rough outline how sendmail 
sequences the milters and rulesets upto the DATA command:

	without FEATURE(`delay_checks')		FEATURE(`delay_checks)
	--------------------------------------------------------------
	call milters for connect              call milters for connect
	check_relay ruleset

	call milters for HELO                 call milters for HELO

	call milters for MAIL                 call milters for MAIL
	check_mail ruleset
	
	call milters for RCPT                 call milters for RCPT
	check_rcpt ruleset
	                                      check_rcpt ruleset
	                                      check_mail ruleset
	                                      check_relay ruleset

So given the results from the log sample, I'd say you don't have 
FEATURE(`delay_checks') enabled, but you do have the milter-sender 
options DeferHeloReject=1 and DeferMailReject=1 set, and you run with 
SpamHaus blacklist (as I do).

NOTE that FEATURE(`delay_checks) executes the 3 rulesets for each RCPT 
and in reverse order after the milter.


> Dec 22 14:26:44 aphrodite sendmail[24262]: NOQUEUE: connect from [222.97.250.150]
> Dec 22 14:26:44 aphrodite sendmail[24262]: iBMMQinF024262: Milter (milter-sender):
init success to negotiate
> Dec 22 14:26:44 aphrodite sendmail[24262]: iBMMQinF024262: Milter (spamassassin):
init success to negotiate
> Dec 22 14:26:44 aphrodite sendmail[24262]: iBMMQinF024262: Milter (mimedefang): init
success to negotiate
> Dec 22 14:26:44 aphrodite sendmail[24262]: iBMMQinF024262: Milter: connect to filters
> Dec 22 14:26:45 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: trying MX 10
'gateway3.delphi.com.' [69.220.142.15] for <220k@delphi.com>
> Dec 22 14:26:45 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: opening SMTP
connection to 69.220.142.15
> Dec 22 14:26:46 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: < 220
gateway3.delphi.com ESMTP Mirapoint 3.4.6-GR; Wed, 22 Dec 2004 17:26:46 -0500 ($
> Dec 22 14:26:46 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: > HELO
aphrodite.jeffandclaire.com^M
> Dec 22 14:26:46 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: < 250
gateway3.delphi.com Hello aphrodite.jeffandclaire.com [68.126.85.58], pleased t$
> Dec 22 14:26:46 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: > MAIL
FROM:<>^M
> Dec 22 14:26:46 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: < 250
<>... Sender ok
> Dec 22 14:26:46 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: > RCPT
TO:<K0220@delphi.com>^M
> Dec 22 14:26:49 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: < 550
<K0220@delphi.com>... No such mailbox
> Dec 22 14:26:49 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: > RSET^M
> Dec 22 14:26:49 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: < 250 Reset
state
> Dec 22 14:26:49 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: > MAIL
FROM:<>^M
> Dec 22 14:26:49 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: < 250
<>... Sender ok
> Dec 22 14:26:49 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: > RCPT
TO:<220k@delphi.com>^M
> Dec 22 14:26:52 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: < 550
<220k@delphi.com>... No such mailbox
> Dec 22 14:26:52 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: > QUIT^M
> Dec 22 14:26:52 aphrodite milter-sender[24125]: 00005 iBMMQinF024262: closing SMTP
connection
> Dec 22 14:26:52 aphrodite spamass-milter[1365]: queueid=iBMMQinF024262
> Dec 22 14:26:52 aphrodite sendmail[24262]: iBMMQinF024262: dns
150.250.97.222.sbl-xbl.spamhaus.org. => 127.0.0.4
> Dec 22 14:26:52 aphrodite sendmail[24262]: iBMMQinF024262: ruleset=check_rcpt,
arg1=<jeff@jeffpowell.com>, relay=[222.97.250.150], reject=553 5.3.0 <jeff@j$
> Dec 22 14:26:53 aphrodite sendmail[24262]: iBMMQinF024262:
from=<220k@delphi.com>, size=0, class=0, nrcpts=0, proto=SMTP, daemon=MTA,
relay=[222.97.250.150]
>  
> In this case, it ultimately did get bounced because the sender was listed in
Spamhaus, but milter-sender should have rejected it first.  I have seen many cases of
this, including viruses, getting through all the way.  Any ideas?
>  
> Jeff
> 


-- 
Anthony C Howe                                 +33 6 11 89 73 78
http://www.snert.com/       ICQ:
7116561         AIM: Sir Wumpus

            "Once...we were here."  - Last of The Mohicans


Lists Index Date Thread Search