Unless you like playing whack-a-mole, you need a smarter hammer, not a
bigger one.
Email peering *IS* a smarter hammer. If all the cluefull email
administrators would set up peering agreements with each other
and exchange contact information, there would be fewer of these
situations. Part of the problem is that there are no agreed
rules of engagement for email abuse issues. By setting up email
peering agreements in advance, we could put those rules of
engagement in place and we could ensure that our email peers
have the *RIGHT* contact information.
Domain registry whois listings and INOC-DBA are not the right
contact information because they are too general. Now, an email
peering agreement could very well specify that certain whois
contact listings should be used as a second resort and that
agreement would make them the right contact info. Also, the
email peering agreement could specify that INOC-DBA phone number
ASNUM*999 is the number one choice of contact method and then it
would become the right way to contact email peers.
The fundamental problem is not that there isn't technology in
place to solve these problems; it is that there aren't *AGREEMENTS*
in place to solve these problems. Organizations like CAUCE are
happy to just bitch and moan instead of working to bring all
email operators together to set up working agreements for
*MANAGING* the email abuse problems instead of always letting
the abusers take the first steps and drive the whole issue.
In fact, given the past experience with two joint ARIN/NANOG meetings
could the best way forward be to have more joint meetings that
combine a NANOG meeting with some other non-BGP/routing operational
forum? Perhaps something jointly with a security organization
like CIS? http://www.cisecurity.org/
There's a lot more people doing SMTP than doing BGP. Also, AS1312 (us) and our
related routing swamp does BGP peering with less than a dozen peers, but we end
up talking SMTP to a good chunk of the world. I've got one machine that all by
itself talked to 2,615 hosts in 1,612 second-level domains yesterday.
Unless you're advocating a return to the X.400-style ADMD/PRMD stuff,
this really is a non-starter. I don't have time to set up 1,600+ peering agreements,
and possibly have to set up more just because somebody subscribes to a mailing list
(either somebody elsewhere subscribes to ours, or one of my users subscribes elsewhere).
And history has passed its own verdict on ADMD/PRMD.
Part of the problem is that there are no agreed rules of engagement
for email abuse issues. By setting up email peering agreements in
advance, we could put those rules of engagement in place and we
could ensure that our email peers have the *RIGHT* contact
information.
Agreed. One of the things that I would really like the new improved
ASRG to work on is email abuse report standards. Just a standard
abuse report format in XML that makes it possible to tell reliably
who's complaining, what they're complaining about, any why they think
you're responsible (IP address, URL, whatever) would be a big step
forward.
Over at APRICOT, CAUCE Asia Pacific (apcauce) is holding an antispam tutorial and conference track, which will also double as an apcauce meeting.
APCAUCE (http://www.apcauce.org) holds tutorials / meetings twice a year, at apricot (Feb) and apan (August) events.
In fact, given the past experience with two joint ARIN/NANOG meetings
could the best way forward be to have more joint meetings that
combine a NANOG meeting with some other non-BGP/routing operational
forum? Perhaps something jointly with a security organization
like CIS? http://www.cisecurity.org/
That, too, would be a good idea. Once some people realize that routers / bgp issues are not the only operational content in existence.
>Unless you like playing whack-a-mole, you need a smarter hammer, not a
>bigger one.
Email peering *IS* a smarter hammer. If all the cluefull email
administrators would set up peering agreements with each other
and exchange contact information, there would be fewer of these
situations.
I think thats a great idea
Domain registry whois listings and INOC-DBA are not the right
contact information because they are too general.
General yes, but calling someone on INOC-DBA will get you a clueful soul.
