[Top] [All Lists]

Re: Bounce/System Notification Address Verification

2005-06-30 12:58:22

On Thu June 30 2005 15:02, Hector Santos wrote:

This is why a good second RCPT TO: check on a random address has alot of
value because it help detect which of the above the host is operation.

A check on a *random* mailbox has zero value because a client cannot
tell whether a supposedly-random local part is or is not a valid
mailbox on a remote system.  Nor can it tell if a supposedly-random
domain name is also legitimately handled by the same MX host, at least
not without doing a separate MX lookup or having equivalent out-of-band

Now, some *specific* (i.e. not at all random) checks might yield some
o permanent failure to accept "postmaster" for the domain indicates RFC
  ignorance (no further tests are likely to be useful)
o rejection of a syntactically-invalid mailbox is probably a good sign;
  conversely acceptance may simply mean that there's a catch-all
o acceptance for a mailbox having a domain known not to be handled by
  the MX may indicate an open relay; but open relays are not necessarily
  always evil
o rejection of a specific mailbox which appears in a specific message
  reverse path for which the host is a cognizant MX might indicate a
  problem -- but the problem might be in the site's configuration

Systems which accept random address (non local domains) are not part of the
solution. They are all intent and purpose open relays

Not necessarily; some hosts serve as MX for multiple domains -- that
is perfectly legitimate.

and will most likely 
get black listed in some DNS RBL site.

Considering that sites which operate no mail service at all tend to
get listed on blacklists, that doesn't say much about the sites; it
says something about blacklists, but that's another matter.