SPAM Level Status - And why not stop the peering with lame ISPs

Alain Hebert ahebert at
Fri Feb 10 04:21:49 UTC 2006


    Yes, those are already in place and do a really good job (about 40% 
from the daily stats).

    Another 40% get caught by razor, pyzor, our own local spam election 
database and spamassassin.  (less than 1% are viruses)

    Its the other 20% which is buggin the hell of our clients...  
(Mostly New spam format and the dynamic spam with generated images)


    I think its more a responsability problem than a technology one.

    All our clients sign a U]sage A]cceptable U]se P]olicy.  Anybody 
caught spamming, spam advertising, warezing, illegal downloaded (when 
BayTSP notify us) get a $500 CDN fine (about $1US) or get disconnected.  
So we take it seriously (and we applied it at least 15 times last year).

    Most ARIN ISP's also take it somewhat seriously (legal issues and 
such)...  Except for those big ones, big lawyers thrump reality/truth 

In summary for now:

    The situation is pretty much statu-quo.

Suresh Ramasubramanian wrote:

>On 2/10/06, Alain Hebert <ahebert at> wrote:
>>    For APNIC, we also includes all their peers up-to (if possible) to a
>>ARIN one.  But we only do that on extreme case of network flooding.
>>    (No sense on wasting operator time on spam related incidents)
>I agree you have a problem there - but try using something like
>'s sbl and xbl first.  And then a few other well chosen
>blocklists (not the "block all traffic from a country" variety at all)
>You wont get any productive results from blocking apnic space the way you do.

Alain Hebert                                ahebert at   
PubNIX Inc.        
P.O. Box 175       Beaconsfield, Quebec     H9W 5T7	
tel 514-990-5911    fax 514-990-9443

More information about the NANOG mailing list