filtering by header without spamassassin
Hi all,
Can e-mail be rejected on channel by specific header without usage of spamassassin?
Have antivirus firewall appliance, Messaging 6.2 with site-wide spamassassin: firewall can either delete virus in e-mail (actually it replaces it with xxxxxxx.txt) or add SMTP header X-Virus (value 1). MTA add ***SPAM*** to subject and store e-mail with spamassassin result higher than 5 into SPAM folder. At moment have option to:
1)Tag viruses with X-Virus, configure spamassasin to score it. This will move virus e-mails to users spam folder (don抰 like)
2)Delete viruses on firewall, and let users inboxes filled with virus-left-overs (spamassassin can be configured to target xxxxxxx.txt files but cant detect file name inside archives which preserve their original names) (don抰 like also)
Also tried to implement per-user spamassassin based on libspam.so . Everything works, except there is error in webmail (when you access filters), something like 揅ant display rule? Rule was added through ldap and works! It can drop viruses, put spam with score 20 directly in trash and score result in range 5-15 into spam folders, all per user customizable. Question for all: if you have this configuration (Messaging 6.2, libspam, LDAP per user), how did you resolve user interaction? We planed to build separate web site where users can change score, but cant find way to avoid webmail error or hide that rule from displaying in filters?Any advice would be apritiated.
We try to simplify configuration and administration. To my knowledge using aliasdetourhost assumes that we need to have additional MTA as base for spamassassin (Please, correct me if I am wrong?).
What is recommended way for highly customizable per-user antispam solution based on spammassassin for 100.000 users? Aliasdetour or libspam?
Gljiva
