Solved

Fine-tune SpamAssassin

Posted on 2004-07-30
7
799 Views
Last Modified: 2011-09-20
HI
I would like to know some tips on fine-tuing SpamAsassin / MailScanner to block more spam.

I've used pretty much "out the box" settings up to now, and it only marks all the messages as spam. I activated SBL+XBL checks yet, I still receive e-mail from such sources, although I have the setting to mark such messages as "High Scoring" and have "delete" as the action for high scoring spam...

Also, I noticed in the config that you can quarantine spam and virus infected messages, however, how do users access these for revision?

0
Comment
Question by:psimation
  • 3
7 Comments
 
LVL 17

Accepted Solution

by:
owensleftfoot earned 63 total points
ID: 11681724
The best way to teach spamassassin is to use sa-learn. Any time spam slips through, get your users to forward it to a local user (on the linux box) such as the user spam. (You will obviously have to create this account). Then use sa-learn to scan the mbox of that user (spam).
http://spamassassin.apache.org/doc/sa-learn.html will show you how.

For the second point, you can setup both mailscanner & spamassassin to send an email to a user who has had a mail blocked because of spam or virus flags with details about how to retrieve the original message.)
Rtfm paul :) ( I think I remember your name is paul? If not, I apologise. You havent been around in a while.)
0
 
LVL 20

Expert Comment

by:Gns
ID: 11692913
I' don't use bayes, and frankly am not sure of it's virtues. It makes it less easy to answer questions like "why is this spam"... :-).
But just to show that I'm not consequent in my thinking... I _will_ swear by the effectiveness of Razor (http://razor.sf.net), Pyzor (pyzor.sf.net) and DCC (http://www.rhyolite.com/anti-spam/dcc/) ... And some "intelligent" whitelisting;).
I also use the excellent MailWatch (http://mailwatch.sourceforge.net/) to visualize the scanning process... Makes it real easy to find problems with the current config ... and fix them.

My guess about the RBLs would be that you don't get enough hits (or even test against enough RBLs to be able) to reach the "High scoring spam" level. ISTR you need 3 (or was it 5) "hits" to mark a message as highscoring. Thus you'd just invoke the usual spam action, which likely is "deliver".

-- Glenn
0
 
LVL 17

Author Comment

by:psimation
ID: 12419872
Hi Guys
I'm still battling with this darn MailScanner.

I changed the required SBL/XBL list hits to 2 , then made it so my high scoring spam score is 6, and made it so that if the message hits at least 2 SBL /XBL lists, it would be marked as high scoring, ie, 6. Then I have the action for messages of more than 6 set to delete.

However, I still get messages that clearly stats in the headers that it has hit at least 2 SBL/XBL lists, AND it gave the message a spam score in total of 7, even though the action for >6 = delete...

Am I using sane values for high scoring spam?

0
 
LVL 20

Expert Comment

by:Gns
ID: 12422845
Eh, do these messages really get _delivered_? Or are you using MailWatch (or similar) to look at 'em? If the latter, the HS spamaction (noted a bit down in the details page) should be delete... making the database log entry all that remains:-).

-- Glenn
0
 
LVL 20

Assisted Solution

by:Gns
Gns earned 62 total points
ID: 12422973
.... And "sane" values for HS spam is very dependant on how well it (spamassassin mostly) behaves, and the "Real Mails" looks like... If I we're to go as low as 6 for HS... I'd be blocking some real mails (which is a nono for my organization (.gov-ish, with a need to be "available to the public". Sigh. This make me only able to store HS spam, so that a person actually looks through the quarantined stuff before deletion. Guess who get's to do the looking:-).
This works for me since we only get about 1500-2000 mails a day (with HS spam at about 1-2%).

My advice would be to play it cool for a while (letting more through than you like) and try to establish working black/whitelists...

I've also reconsidered my position on bayes... I'm using (and loving) it... Only tricky part with it is that one needs be able to get the pure unchanged message to train it properly (when missdetections occur), but that can be handled by some form of IMAP folder where the users can dump messages (one for ham, one for spam).

-- Glenn
0

Featured Post

Master Your Team's Linux and Cloud Stack!

The average business loses $13.5M per year to ineffective training (per 1,000 employees). Keep ahead of the competition and combine in-person quality with online cost and flexibility by training with Linux Academy.

Question has a verified solution.

If you are experiencing a similar issue, please ask a related question

Suggested Solutions

Network Interface Card (NIC) bonding, also known as link aggregation, NIC teaming and trunking, is an important concept to understand and implement in any environment where high availability is of concern. Using this feature, a server administrator …
Little introduction about CP: CP is a command on linux that use to copy files and folder from one location to another location. Example usage of CP as follow: cp /myfoder /pathto/destination/folder/ cp abc.tar.gz /pathto/destination/folder/ab…
Learn several ways to interact with files and get file information from the bash shell. ls lists the contents of a directory: Using the -a flag displays hidden files: Using the -l flag formats the output in a long list: The file command gives us mor…
Learn how to navigate the file tree with the shell. Use pwd to print the current working directory: Use ls to list a directory's contents: Use cd to change to a new directory: Use wildcards instead of typing out long directory names: Use ../ to move…

838 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question