I'm having a heck of a time getting whitelist_from_rcvd to work AT ALL since upgrading to 2.63 . I have read the threads about trusted networks, etc., but nothing seems to work. (To be honest, I don't see the logic behind it only working on trusted networks... it may foil some spoofs but it makes administration a nightmare.)
That said... I can't even get outgoing mail from our very own 192.168.1.XXX network to trigger a whitelist_from_rcvd rule, even when I explicitly declare that network trusted. I have had to resort to local meta rulesets that check the From and Received lines for the desired combination of strings
That right there is your problem most likely... If you are using a NATed mailserver you MUST manualy declare trusted_networks. Otherwise SA mis-understands the mail headers and decides your SMTP gateway is a server outside your network.
Try forcing a trusted_networks statement that contains _only_ your mailserver. Do not include your hosts. Do not include any outside ISPs. Just include your servers.
