Milter-greylist

Hi,

Am Sonntag, den 12.03.2017, 11:54 -0400 schrieb Greg Troxel 
gdt@... [milter-greylist]:
"Thomas Cameron thomas.cameron@..." writes:
> >
> > I had a mail conversation with someone today. They use some Amazon
> > hosted mail service. The first email came from [sender1] and was relayed
> > through ec2-54-153-11-11.us-west-1.compute.amazonaws.com. The second
> > mail came from [sender1] but from
> > ec2-54-153-22-22.us-west-1.compute.amazonaws.com, and so on. Every new
> > e-mail wound up coming through a different relay on Amazon.
> >
> > What winds up happening is, EVERY new e-mail is delayed, often times
> > several hours because they don't retry in a timely fashion.

I've seen that too, especially with Amazon emails.

[This list has bad behavior with Reply-To and rewriting the sender....]
> 
> A few ideas:
> 
> * There's a notion of whitelisting address ranges that are inhabited by
>   this kind of distributed retrying.
> 
> * In an age where greylisting is normal, it's buggy of a sender not to
>   retry from the same address.  Good luck with that approach :-)

I like that idea :D

* Perhaps milter-greylist could by default or could be configured to
>   consider hosts in the same /24 (not enough for your case) or /16 to 
> be
>   the same.

check subnetmatch option, e.g.:

# all addresses within the same class C network are considered the same
subnetmatch /24

Ciao
Marcus