Postby adhodgson » Sun Jun 03, 2007 9:07 am
Hi,
The problem with DSNs is that they stop when the mail encounters a SMTP server which doesn't support DSNs.
If you send a message from a non-scalix account, asking for a DSN (say using Outlook), and all mail servers between you and the Scalix server supports DSN commands, you end up receiving a message from the Scalix server saying the mail has been delivered. This comes from <postmaster@machinename.domain.com>.
If you send from a Scalix client to an external user, then depending on how the Scalix server sends mail out, several things could happen:
1. The DSN will be handled by the relevant server if all the servers in the chain support DSNs.
2. If Sendmail on the Scalix box sends messages directly to a server that doesn't support the DSN command, or if it sends a message via a smarthost that doesn't support a DSN command, then it will generate a message relayed message, which usually comes from <mailer-daemon@machine.domain.com>, and has gone through Sendmail.
3. If another machine further down the mail loop doesn't support DSNs, then the machine that is sendingg the message to the non-DSN aware MTA will generate a relayed success message. For example, where our Scalix box is located, it sends messages through a spam scanner, which quite often generates relayed DSNs for us.
So it is probably a good idea to look at how which users are requesting the DSN, where they are requesting it (i.e, are they senders sending to a user on Scalix, or users on Scalix sending externally), and to which addresses are they sending the mail to. Looking at the headers will give you an indication of where the DSN was generated also.
Thanks.
Andrew.