On Wed, Dec 05, 2001 at 02:45:32PM +0200, Sharun@??? wrote:
> Okay, I solve this problem with couple of external scripts. Interesting ?
> 2 hours of work = 250 hosts in my RBL :))
Fair, but how long did the script take to decide whether or not the host
was relaying? It seems fair to say that. Also, I believe your script is
badly written:
2001-12-05 12:52:55 GMT: reject: command host=storage.ukr.net \
from=mbm@??? cmd="RCPT TO:<mail@???>"
resp="550 relaying to <mail@???> prohibited by administrator" \
rcpt-normal-550=mail@??? irritated=24004ms
I don't believe "storage.ukr.net" has any business masquerading as me.
If you want to do a relay test, do it from domains you own. This comes into
my point 3 - you'll get it wrong.
Which reminds me of my other point
6) there is no *one* way of relaying, different setups have different bugs
which, in some cases, will allow you to use, for example the percent
hack, or perhaps they'll do source-routing of mails, so the number of
tests you have to do, and to understand is relatively high.
Remembering, of course, to distinguish between straight relays and
multihop relays.
MBM
--
Matthew Byng-Maddick <mbm@???> http://colondot.net/
