You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@spamassassin.apache.org by sb ch <ch...@hotmail.com> on 2004/06/25 10:46:24 UTC

curious problem about spam score...

Hello, all.

I'm a beginner of the spamassaaain.
and I have read documents and FAQ, But some problem which I can't 
understand.
Please help me.


I have operated spamassassin 2.63 at Redhat linux 6.2.
But as below, this mail is "hits=-0.8" and "required=5.0". 
But this mail is classified as SPAM, why?
this list is not blacklist.

*****E-mail header ***** 
>From nobody  Fri Jun 25 14:41:14 2004
Date: Fri, 25 Jun 2004 14:41:14 +0900
To: master@bbb.com 
Subject: ***SPAM***  test
From: aaaa@aaa.com
Content-Type: text/html; charset=euc-kr
X-Spam-Level:
X-Spam-Checker-Version: SpamAssassin 2.63 (2004-01-11)
X-Spam-Status: No, hits=-0.8 required=5.0 tests=BAYES_30,HTML_MESSAGE,
        MIME_HEADER_CTYPE_ONLY autolearn=ham version=2.63


local.cf file is below. 

required_hits           5
rewrite_subject         1
subject_tag           ***SPAM***
report_header           0
use_terse_report       0
defang_mime            0
report_safe               0
use_bayes               1
bayes_path /home/spam/.spamassassin/bayes
auto_whitelist_path /home/spam/.spamassassin/auto-whitelist
bayes_file_mode 777
auto_whitelist_file_mode 777
skip_rbl_checks         1
auto_learn              1
ok_languages            all
ok_locales              all
allow_user_rules        1

_________________________________________________________________
상큼한 만남과 따뜻한 공동체 생활...  지금 MSN 커뮤니티에서 시작하세요!   
http://groups.msn.com/?pgmarket=ko-kr  


Re: curious problem about spam score...

Posted by Loren Wilton <lw...@earthlink.net>.
This is most strange.  The only thought that comes to mind is that somehow
this message has been through SA twice.  The first time it was spam and got
the subject markup, and probably some X-Spam lines showing why it was spam.
Probably also put the real spam in an attachment.  The second time through
it wasn't recognized as spam, so the original X-Spam markup was replaced,
but the subject was not rewritten.  Leaving it marked as spam from the first
time.

        Loren

> I have operated spamassassin 2.63 at Redhat linux 6.2.
> But as below, this mail is "hits=-0.8" and "required=5.0".
> But this mail is classified as SPAM, why?
> this list is not blacklist.
>
> *****E-mail header *****
> >From nobody  Fri Jun 25 14:41:14 2004
> Date: Fri, 25 Jun 2004 14:41:14 +0900
> To: master@bbb.com
> Subject: ***SPAM***  test
> From: aaaa@aaa.com
> Content-Type: text/html; charset=euc-kr
> X-Spam-Level:
> X-Spam-Checker-Version: SpamAssassin 2.63 (2004-01-11)
> X-Spam-Status: No, hits=-0.8 required=5.0 tests=BAYES_30,HTML_MESSAGE,
>         MIME_HEADER_CTYPE_ONLY autolearn=ham version=2.63
>
>
> local.cf file is below.
>
> required_hits           5
> rewrite_subject         1
> subject_tag           ***SPAM***
> report_header           0
> use_terse_report       0
> defang_mime            0
> report_safe               0
> use_bayes               1
> bayes_path /home/spam/.spamassassin/bayes
> auto_whitelist_path /home/spam/.spamassassin/auto-whitelist
> bayes_file_mode 777
> auto_whitelist_file_mode 777
> skip_rbl_checks         1
> auto_learn              1
> ok_languages            all
> ok_locales              all
> allow_user_rules        1
>
> _________________________________________________________________
> 상큼한 만남과 따뜻한 공동체 생활...  지금 MSN 커뮤니티에서 시작하세요!
> http://groups.msn.com/?pgmarket=ko-kr
>


Re: Question about spoofing

Posted by Mariano Absatz <sp...@lists.com.ar>.
Hi Nelly,

you shouldn't whitelist your users's e-mail addresses but the IP addresses 
from where your users send mail (I think it's done with whitelist_from_rcvd, 
I don't use it since I use MailScanner's whitelist).

As you've already seen, mail from spoofing using valid internal e-mail 
addresses is _very_ common in spam.


El 25 Jun 2004 a las 10:37, Nelly Yusupova escribió:

> Hello,
> 
> I'm running SA version of 2.63 with procmail on  Solaris 2.6.  The spam
> software works great and blocks most of the spam.  Mail Spoofing is still a
> big problem though.
> 
> I added all of the internal valid email address to the whitelist to make
> sure that all of the people at work can send EVERYTHING to each other.
> However this also lets all of the spam that uses mail spoofing through.
> 
> Here is an example of a header:
> 
> Return-Path: <km...@cgim.com>
> Received: from cm218-254-210-162.hkcable.com.hk
> (cm218-254-210-162.hkcable.com.hk [218.254.210.162])
> 	by cgim.com (8.12.10/8.12.10) with SMTP id i5P3vaq2018122
> 	for <km...@cgim.com>; Thu, 24 Jun 2004 23:57:38 -0400 (EDT)
> Message-Id: <20...@cgim.com>
> X-Message-Info: 39J6Wxmdof660ymwWyF75IZ4djSEptXH6
> Received: from charitable.kmk@cgim.com
> (nt43575.l09393228.j-bo.naq.kmk@cgim.com [77.150.188.178])
> 	by socialemphases.kmk@cgim.com
> 	id YWY1572; Sun, 27 Jun 2004 11:12:23 +0600
> 	[bayonetHost SMTP Relay 79.931]
> Reply-To: "Otis Coulter" <km...@cgim.com>
> From: "Otis Coulter" <km...@cgim.com>
> To: "Kmk" <km...@cgim.com>
> Subject: Kmk VAL1UM, XNAX,LEV1TRA..S0MA..MUCH MORE...
> Date: Sun, 27 Jun 2004 02:12:23 -0300
> MIME-Version: 1.0
> Content-Type: multipart/alternative;
> 	boundary="--=====08032671409=_"
> X-Spam-Checker-Version: SpamAssassin 2.63 (2004-01-11) on io.cgim.com
> X-Spam-Status: No, hits=-70.8 required=2.5 tests=BAYES_99,DRUGS_ANXIETY,
> 	DRUGS_ANXIETY_EREC,DRUGS_ANXIETY_OBFU,DRUGS_ERECTILE,DRUGS_MUSCLE,
> 	HTML_50_60,HTML_IMAGE_RATIO_04,HTML_MESSAGE,J_CHICKENPOX_22,
> 	J_CHICKENPOX_32,LOC_TINY_FONT_1,MIME_HTML_ONLY,MIME_HTML_ONLY_MULTI,
> 	MSGID_FROM_MTA_HEADER,RM_rbt_Font0Pt,TW_DX,USER_IN_WHITELIST
> 	autolearn=no version=2.63
> X-Spam-Level:
> Status:
> 
> Is there a way to stop this spam even though the user is in the whitelist?
> 
> In this header the user uses a different name associated with and email
> address.  Is there a way to specify a name with and email so that the
> incoming mail only goes through if both the name AND the email address
> match?
> 
> From: "Otis Coulter" <km...@cgim.com>
> 
> Thanks in advance for all your help.
> 
> Nelly.
> 


--
Mariano Absatz
El Baby
----------------------------------------------------------
Love is the answer, but while you're waiting for the
answer, sex raises some pretty interesting questions.
        -- Woody Allen




Re: Question about spoofing

Posted by Matt Kettler <mk...@evi-inc.com>.
At 10:37 AM 6/25/2004, Nelly Yusupova wrote:
>Is there a way to stop this spam even though the user is in the whitelist?

Don't whitelist your local users, or use whitelist_from_rcvd instead of 
whitelist_from.





Re: Question about spoofing

Posted by Duncan Hill <sa...@nacnud.force9.co.uk>.
On Friday 25 June 2004 15:37, Nelly Yusupova might have typed:
> Hello,

> I added all of the internal valid email address to the whitelist to make
> sure that all of the people at work can send EVERYTHING to each other.
> However this also lets all of the spam that uses mail spoofing through.

> Is there a way to stop this spam even though the user is in the whitelist?
>

Why are you running internal mail through filters that only need to apply to 
external mail?  Whitelisting your addresses is a bad idea as you've now found 
out, as spammers have no compunction about sending as you.

Re-configure your network to apply SA only on mail from the outside world in, 
and implement things like DNSBL checks against Spamhaus at the SMTP level.  
Internal mail should probably never be scanned for spam, unless you have spam 
problems with internal mail :)

Question about spoofing

Posted by Nelly Yusupova <ne...@cgim.com>.
Hello,

I'm running SA version of 2.63 with procmail on  Solaris 2.6.  The spam
software works great and blocks most of the spam.  Mail Spoofing is still a
big problem though.

I added all of the internal valid email address to the whitelist to make
sure that all of the people at work can send EVERYTHING to each other.
However this also lets all of the spam that uses mail spoofing through.

Here is an example of a header:

Return-Path: <km...@cgim.com>
Received: from cm218-254-210-162.hkcable.com.hk
(cm218-254-210-162.hkcable.com.hk [218.254.210.162])
	by cgim.com (8.12.10/8.12.10) with SMTP id i5P3vaq2018122
	for <km...@cgim.com>; Thu, 24 Jun 2004 23:57:38 -0400 (EDT)
Message-Id: <20...@cgim.com>
X-Message-Info: 39J6Wxmdof660ymwWyF75IZ4djSEptXH6
Received: from charitable.kmk@cgim.com
(nt43575.l09393228.j-bo.naq.kmk@cgim.com [77.150.188.178])
	by socialemphases.kmk@cgim.com
	id YWY1572; Sun, 27 Jun 2004 11:12:23 +0600
	[bayonetHost SMTP Relay 79.931]
Reply-To: "Otis Coulter" <km...@cgim.com>
From: "Otis Coulter" <km...@cgim.com>
To: "Kmk" <km...@cgim.com>
Subject: Kmk VAL1UM, XNAX,LEV1TRA..S0MA..MUCH MORE...
Date: Sun, 27 Jun 2004 02:12:23 -0300
MIME-Version: 1.0
Content-Type: multipart/alternative;
	boundary="--=====08032671409=_"
X-Spam-Checker-Version: SpamAssassin 2.63 (2004-01-11) on io.cgim.com
X-Spam-Status: No, hits=-70.8 required=2.5 tests=BAYES_99,DRUGS_ANXIETY,
	DRUGS_ANXIETY_EREC,DRUGS_ANXIETY_OBFU,DRUGS_ERECTILE,DRUGS_MUSCLE,
	HTML_50_60,HTML_IMAGE_RATIO_04,HTML_MESSAGE,J_CHICKENPOX_22,
	J_CHICKENPOX_32,LOC_TINY_FONT_1,MIME_HTML_ONLY,MIME_HTML_ONLY_MULTI,
	MSGID_FROM_MTA_HEADER,RM_rbt_Font0Pt,TW_DX,USER_IN_WHITELIST
	autolearn=no version=2.63
X-Spam-Level:
Status:

Is there a way to stop this spam even though the user is in the whitelist?

In this header the user uses a different name associated with and email
address.  Is there a way to specify a name with and email so that the
incoming mail only goes through if both the name AND the email address
match?

From: "Otis Coulter" <km...@cgim.com>

Thanks in advance for all your help.

Nelly.



Re: curious problem about spam score...

Posted by Matt Kettler <mk...@comcast.net>.
At 05:46 PM 6/25/04 +0900, sb ch wrote:
>I have operated spamassassin 2.63 at Redhat linux 6.2.
>But as below, this mail is "hits=-0.8" and "required=5.0". But this mail 
>is classified as SPAM, why?
>this list is not blacklist.
>
>*****E-mail header ***** From nobody  Fri Jun 25 14:41:14 2004
>Date: Fri, 25 Jun 2004 14:41:14 +0900
>To: master@bbb.com Subject: ***SPAM***  test
>From: aaaa@aaa.com
>Content-Type: text/html; charset=euc-kr
>X-Spam-Level:
>X-Spam-Checker-Version: SpamAssassin 2.63 (2004-01-11)
>X-Spam-Status: No, hits=-0.8 required=5.0 tests=BAYES_30,HTML_MESSAGE,
>        MIME_HEADER_CTYPE_ONLY autolearn=ham version=2.63

The most likely case is that this message got scanned twice, and the SPAM 
header was added the first time it was scanned. The second time through the 
X-Spam-Status was over-written.

Does the tagged message have a body report? Do the listed hits differ from 
X-Spam-Status?