In Plesk versions earlier than 11.5, the outgoing mail of all customers was sent from a single IP address (defined by the mail server configuration). Thus, if one of the customers became blacklisted for sending spam, other customers were automatically blacklisted too since they used the same IP address. Also, if a customer had several IP addresses, and the address for outgoing mail did not match the address of the domain, the customer run a risk to be blacklisted as well.

In Plesk 11.5, the problem of domain and mail addresses is resolved, and Postfix mail server uses customers’ IP addresses for sending mail if possible. This targets all outgoing mail of the Plesk mail server sent by PHP mail(), sendmail, an SMTP script or client. However, if the following conditions are true, your system can send mail from different IP addresses:

• You have Postfix 2.7
• Plesk is configured to support IPv6
• A subscription has only a single IPv4 or IPv6

The outgoing mail for such subscriptions can be sent from either the customer’s IP address or an IP address of one type (IPv6 for IPv4 and vice versa) specified in the mail server configuration.

Another part of this feature is the sender’s address validation: The system validates the MAIL FROM header for authenticated users and corrects the header if needed. However, if the mail is sent without authentication, for example, from 127.0.0.1, through the local sendmail, or a sender is in the white list, the system trusts the MAIL FROM header.

And in case you need to send all mail from one specific IP address, you can specify one IPv4 or one IPv6 address from which all mail will be sent. All corresponding settings are available in the Plesk GUI in Plesk for Linux with Postfix as a mail server. For details, see the Plesk Administrator’s Guide, Configuring Server-Wide Mail Settings.

Requirements

This feature is supported on all operating systems that have Postfix 2.7 or later. They are as follows: