Here are a few things I am doing with QS & SA, Don't know if they are applicable to your situation.
1) In my case the server runs in a small company and almost all of the emails are delivered to an Exchange Server. I've setup 2 public folders called SPAM and HAM. The users drag emails in outlook to those folders and there is a fetchmail script which runs every 1/2 hour on the mail server, gathers those emails via IMAP and pipes them to sa-learn. 2) We have a rather low score for spam - 4 points, above that the emails have their subject rewritten to include a SPAM token and then there are rules in Outlook to put those emails into the users' "Deleted Items" folder (incase they still want to look at it). Above 8.5 points the emails are quarantined (I am using the beta version of qs 1.3, by using the ST patch you can delete emails above a certain score). We have some real emails classified as SPAM, but not too much, it's a trade off. 3) Beef up your SA configuration. If you have a moderate email quantity (not an ISP) use RBL and URBL filtering (they work really good) and the razor/dcc filtering. Also make sure you are using the latest release version of SA. 4) You might want to some custom rules, for instance I have a few rules which lower the score for specific tokens found in our company email signatures, so any reply with the original text inside will have less chance to be considered spam. Also if have tokens specific to the business of your company, you can add those manually with a negative score. For instance a cloths retailer might add "Pants" "Shirt" "Blouse" etc, to lower the score by 0.5 points. (works good in my case, not those specific words) 5) Make sure your auto white list is working, it helps greatly to reduce the number of good emails considered spam if you are using a low score. That's all that comes to mind, Please share other ideas -----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On Behalf Of Liutauras Adomaitis Sent: Tuesday, December 06, 2005 13:18 To: qmail-scanner-general@lists.sourceforge.net Subject: [Qmail-scanner-general]Best practice for QS+SpamAssassin hello all, I don't know if this is the right list, but I wonder is there any information about what is best practice in dealing with spam. My problem is that users get a lot of spam despite of QS+SA running on my mail server. I think that I should run sa-learn, but I cannot get those spam messages since they are already delivered to users computers. So how to catch spam and make SA learn? -------------------------------- Liutauras Adomaitis ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click _______________________________________________ Qmail-scanner-general mailing list Qmail-scanner-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/qmail-scanner-general ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=7637&alloc_id=16865&op=click _______________________________________________ Qmail-scanner-general mailing list Qmail-scanner-general@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/qmail-scanner-general
