Sándor Zsolt wrote:
I see, this is a self protection method. It's okay - I think - if the server
is directly on the Internet, but in our case it is behind a firewall what
does a prefiltering and relays the good domained messages to the mail server.
(The firewall can not check wheather the user in the domain exists or not it
simply relays all of them.)
What software is this firewall running? At my previous job, we were
running Sophos PureMessage, which had no native capability to validate
recipients. However, the system was flexible enough that there were two
rather simple methods of introducing recipient validation. Since the
product used sendmail as its MTA, we set up a job to build a full list
of recipients (users and aliases), and create a new virtusertable on the
filtering host. Another option would have been to create a "list" of
addresses in the products UI, and then create a rule that rejects
recipients not in that list.
Whatever you're using almost certainly will allow you to do one of those
two. I never saw a product that wouldn't, when I was evaluating
different solutions.