utf-8 / base 64 encoded subject

GKrem

Member
Apr 10, 2019
13
1
8
61
i get a lot of SPAM with same words in subject.
So i wrote a rule that is blocking them.
It looks like the rule will not act on utf-8 / base 64 encoded subject?
Is this by design or can i change this?
 
Could you please provide an example mail (at least the headers in 'raw' format ), the rule you created and your logs?
 
Hello Stoiko,
i am using this "What" regex:

(?i)(\W|^)(TcOkZGNoZW4=|date|mädchen|frauen|frau|madchen|liebe|dating|sex|sexuell)(\W|$)

You can see, i tested to include a base64 string but i guess this is very sensless :-(
The whole mail is attached, including the header.
What log do you need? the regular syslog does not give info about the spam handling?

The main issue is with the Subject encoded in utf-8?B - is it possible to have the SPAM filter to first decode the field before the regex is executed?
 

Attachments

  • Example-Mail.txt
    15.7 KB · Views: 14
this might help - i wonder how to follow one mail in the pmg log file? The message id seems to change and i can not find any common string there?
Jun 14 06:26:02 pmg postfix/smtpd[120869]: connect from pmg.xxx-xx.local[192.168.0.202]
Jun 14 06:26:02 pmg postfix/smtpd[120869]: AB1FB238E0: client=pmg.xxx-xx.local[192.168.0.202]
Jun 14 06:26:02 pmg postfix/smtpd[120869]: DC71E238E7: client=pmg.xxx-xx.local[192.168.0.202]
Jun 14 06:26:02 pmg postfix/cleanup[120872]: DC71E238E7: message-id=<0a2b62c93df9d3dc3f95dc779296bcbe158cf2@student.tp.edu.sg>
Jun 14 06:26:02 pmg postfix/qmgr[930]: DC71E238E7: from=<1301882E@STUDENT.TP.EDU.SG>, size=14521, nrcpt=1 (queue active)
Jun 14 06:26:03 pmg pmg-smtp-filter[120330]: 238EB62A80DDAEE37E: new mail message-id=<0a2b62c93df9d3dc3f95dc779296bcbe158cf2@student.tp.edu.sg>#012
Jun 14 06:26:03 pmg postfix/smtpd[120869]: 1ED07238EC: client=pmg.xxx-xx.local[192.168.0.202]
Jun 14 06:26:03 pmg postfix/smtpd[120869]: disconnect from pmg.xxx-xx.local[192.168.0.202] ehlo=1 mail=3 rcpt=3 data=3 quit=1 commands=11
Jun 14 06:26:19 pmg pmg-smtp-filter[120330]: 238EB62A80DDAEE37E: SA score=1/5 time=16.580 bayes=undefined autolearn=no autolearn_force=no hits=DKIM_INVALID(0.1),DKIM_SIGNED(0.1),FROM_STARTS_WITH_NUMS(0.553),HTML_MESSAGE(0.001),KAM_DMARC_STATUS(0.01),KAM_GOOGLE_REDIR(0.5),NO_DNS_FOR_FROM(0.379),SPF_PASS(-0.001),T_SCC_BODY_TEXT_LINE(-0.01),T_SPF_HELO_TEMPERROR(0.01)
Jun 14 06:26:19 pmg postfix/smtpd[120897]: connect from localhost.localdomain[127.0.0.1]
Jun 14 06:26:19 pmg postfix/smtpd[120897]: ADC9A238EC: client=localhost.localdomain[127.0.0.1], orig_client=pmg.xxx-xx.local[192.168.0.202]
Jun 14 06:26:19 pmg postfix/cleanup[120872]: ADC9A238EC: message-id=<0a2b62c93df9d3dc3f95dc779296bcbe158cf2@student.tp.edu.sg>
Jun 14 06:26:19 pmg postfix/qmgr[930]: ADC9A238EC: from=<1301882E@STUDENT.TP.EDU.SG>, size=15450, nrcpt=1 (queue active)
Jun 14 06:26:19 pmg postfix/smtpd[120897]: disconnect from localhost.localdomain[127.0.0.1] ehlo=1 xforward=1 mail=1 rcpt=1 data=1 commands=5
Jun 14 06:26:19 pmg pmg-smtp-filter[120330]: 238EB62A80DDAEE37E: accept mail to <t-online@xxxxxxxxx.de> (ADC9A238EC) (rule: default-accept)
Jun 14 06:26:19 pmg pmg-smtp-filter[120330]: 238EB62A80DDAEE37E: processing time: 16.747 seconds (16.58, 0.095, 0)
Jun 14 06:26:19 pmg postfix/lmtp[120877]: DC71E238E7: to=<t-online@xxxxxxxxx.de>, relay=127.0.0.1[127.0.0.1]:10024, delay=17, delays=0.05/0.02/0/17, dsn=2.5.0, status=sent (250 2.5.0 OK (238EB62A80DDAEE37E))
Jun 14 06:26:19 pmg postfix/qmgr[930]: DC71E238E7: removed
Jun 14 06:26:19 pmg postfix/smtp[120898]: ADC9A238EC: to=<t-online@xxxxxxxxx.de>, relay=192.168.0.3[192.168.0.3]:25, delay=0.03, delays=0.01/0/0/0.01, dsn=2.0.0, status=sent (250 Message accepted for delivery)
Jun 14 06:26:19 pmg postfix/qmgr[930]: ADC9A238EC: removed
 
Ok - I think i see the issue - it boils down to the one described in:
https://bugzilla.proxmox.com/show_bug.cgi?id=2057

Currently PMG does not deal well with non-ascii characters (we're planning on fixing this soon) - so the issue with the subject is
Mädchen

The one workaround you can use is replace the Umlauts with '.' (of course that opens things up for false positivies - but I don't thing that you'll miss mails where the subject matches 'm.dchen' if you don't want mails where it contains 'mädchen'

I hope this helps!
 
  • Like
Reactions: GKrem
Hello Stoiko,

i justg want to give you a feedback.

After somed days of silence the spam came back. But now i have the rules activated and this is keeping mail inbox cleaner as ever before.

This is the rules on field subject:

(?i)(\W|^)(TcOkZGNoZW4=|date|m.dchen|frauen|frau|madchen|liebe|dating|sex|sexuell)(\W|$)
(?i)(\W|^)(cannabis|abnehmen|zimt|bauch|gewicht|fett|bauchfett)(\W|$)

Thank yxou for the support and the solution.

Best wishes

Günter
 
  • Like
Reactions: Stoiko Ivanov
Glad that worked as mitigation - the acutally correctly matching non-ascii characters is quite high on our todo list - so soon this should become more straight-forward!
 

About

The Proxmox community has been around for many years and offers help and support for Proxmox VE, Proxmox Backup Server, and Proxmox Mail Gateway.
We think our community is one of the best thanks to people like you!

Get your subscription!

The Proxmox team works very hard to make sure you are running the best software and getting stable updates and security enhancements, as well as quick enterprise support. Tens of thousands of happy customers have a Proxmox subscription. Get yours easily in our online shop.

Buy now!