Oohhhhh.
This really illistruates a lot.

At first glance I thought the /@b\./ from_senders ACL wasn't working, but
realizing a lot of sources START their from address with a b. as well

Ex:
unknown[69.6.27.26]:
from=<[EMAIL PROTECTED]>
to=<[EMAIL PROTECTED]> helo=<03.pn01.com>
unknown[69.6.27.26]:
from=<[EMAIL PROTECTED]>
to=<[EMAIL PROTECTED]> helo=<03.pn01.com>
unknown[69.6.27.26]:
from=<[EMAIL PROTECTED]>
to=<[EMAIL PROTECTED]> helo=<03.pn01.com>
unknown[69.6.27.26]:
from=<[EMAIL PROTECTED]>
to=<[EMAIL PROTECTED]> helo=<03.pn01.com>
unknown[69.6.27.26]:
from=<[EMAIL PROTECTED]>
to=<[EMAIL PROTECTED]> helo=<03.pn01.com>

I'm digging through looking for any possible legit bounce-from address from
opt-in lists like this..
But I'm thinking we need another regexp to catch for this.

Maybe something forcing at least 3 decimals before the @?
I'm scratchy on my regexp's... But something like
/^b\..+\..+\..+/ 

Maybe?

What do you think len?

Want to make sure it wouldn't catch first [EMAIL PROTECTED]
[EMAIL PROTECTED] or something



-----Original Message-----
From: Len Conrad [mailto:[EMAIL PROTECTED] 
Sent: Thursday, August 28, 2003 7:14 PM
To: [EMAIL PROTECTED]
Subject: [IMGate] Re: want to see how good, or bad, it really is?



meet me at the bottom
1. egrep -i "4tuple" /var/log/maillog | awk '{print $10" "$17" "$18" "$20}' 
| sort -t[ -k2 | less

... sort by the from= field, but on the sender.domain.

you can also change the order of the $10, $17, $18, $20 so the 2nd, 3rd, or 
4th fields are printed in first column, nice and straight, easier to
analyze.

etc, etc.

Play with it, it will really pay off.

I've already located a bunch of .br, .cl, .py, .ar subscriber nets with PTR 
hostnames spamming like hell to add to my mta_clients_subscriber.regexp.

Len



Reply via email to