Hi Benoit, 

I assume you already checked M3AAWG's BCPs? 
https://www.m3aawg.org/published-documents
I'm not familiar with _all_ of them, but that's a good source of ideas anyway. 

Cheers, 
--
Benjamin

-----Original Message-----
From: mailop <mailop-boun...@mailop.org> On Behalf Of Benoit Panizzon
Sent: Monday, 1 October, 2018 11:23
Cc: mailop@mailop.org
Subject: [mailop] How to find 'low flying' spamers? (Re: outlook.com blocking 
reason: S3150 "network is on our block list")

Hi List

Well thank you for all the hints. I also (thanks to Al) found out, that you 
need to set the browser language to english, to get to the propper help page 
where the delisting request form can be found. With german, you're lost :-)

But I would like to use that topic on a discussion about what to do to mitigate 
abuse of an ISP email platform.

I believe we do much to mitigate abuse from our email platform, but perhaps, 
there is more that can be done. So tell me how you 'do' it and where you see 
potential for improvement.

First let me explain what we do...

We are a classical end user ISP. Our customers are private households with an 
internet connection and some companies. We encourage companies to use an own 
mailserver. We do not offer any 'relay' services to them, so we won't get the 
crap some hacked company servers produce via our mail platform (of course our 
abuse desk also deals with them, to notify those customers in our AS).

So we mainly get the usual problems with customers who hand out their email 
credentials in reply to phishing emails or get trojans who steal them from 
their computers.

To mitigate those problems we have implemented those mechanisms:

* Require minimal password strength.
* Require SMTP Authentication.
* Keep cache history from which IP an SMTP Authentication occurred.
* If count(IP) in delta time > IPlimit block account and require
  password change.
* If count(geoIP) in delta time > Geolimit block account and require
  password change.

Some exception for google services and local mobile networks NAT which use a 
different ip for each SMTP connections.

* If count(recipients) in delta time > RecipientLimit - tempfail and
  notify postmaster to check manually.

We also use account encapsulated SRS for forwarded emails, so we can detect 
bounces from the far side and deactivate bouncing forwards to limit backscatter 
to a minimum.

Of course, this fails if the phished account keeps below our radar by using a 
single ip address and sending emails slowly to keep below the tempfail 
threshold.

In this case we rely on feedback loops to 'react' and also sometimes on 
customers reporting the bounces their phished account produces as spam :-)

What else could we do?

Mit freundlichen Grüssen

-Benoît Panizzon-
-- 
I m p r o W a r e   A G    -    Leiter Commerce Kunden
______________________________________________________

Zurlindenstrasse 29             Tel  +41 61 826 93 00
CH-4133 Pratteln                Fax  +41 61 826 93 01
Schweiz                         Web  http://www.imp.ch
______________________________________________________

_______________________________________________
mailop mailing list
mailop@mailop.org
https://chilli.nosignal.org/cgi-bin/mailman/listinfo/mailop
_______________________________________________
mailop mailing list
mailop@mailop.org
https://chilli.nosignal.org/cgi-bin/mailman/listinfo/mailop

Reply via email to