Hi folks-
The company that I work for is a Windows Environment site, with a couple of Linux servers here and there for emails and database.
Currently, I am the UNIX admin trying to set up spam filtering on the front end of our incoming email environment.
What I would like is to be able to set up these two mailboxes on the Linux server, and send spam and ham emails from my Outlook to there. The Linux server would then run sa-learn automatically to create its own spam filter.
This is our current email flow
(INternet)
|
|
\/
(Linux Mail Server) [With Postfix+SpamAssassin]
|
|
\/
(Exchange 2003) [All Users mailboxes lie here]
End users are reporting that there are getting bombarded with spam and that SpamAssassin and Exchange's Intelligent Message Filtering is not doing their job. I checked SpamAssassin and did a dump for the Bayesian filter but nothing came up.
I am following this webpage here:
http://gtmp.org/pub/sa-postfix.en.html . I basically setup an alias address on my email said that will forward any emails like mr.spam@somecompany.com back to the Linux Mail Server. In the LMS, I followed the webpage and made an alias for spam and ham, changed the transport file to forward these mails into whatever I specified in the master.cf file.
However, it is not working...
Currently getting this error:
Oct 16 14:48:10 mail postfix/qmgr[32342]: warning: transport sa-spam failure -- see a previous warning/fatal/panic logfile record for the problem description
Oct 16 14:48:10 mail postfix/qmgr[32342]: warning: premature end-of-input on private/sa-spam socket while reading input attribute name
Oct 16 14:48:10 mail postfix/qmgr[32342]: warning: private/sa-spam socket: malformed response
I'm not sure what the procedure is, but is there anything I am not doing right?
Thanks in advance
Wilson
Start Free Trial
