Hi ! We are a little internet-provider and I'm the postmaster of it.
We have 7 mailservers, 2 incoming (MX) two spoolers, and 2 outgoing for our customers. We have 1 Spamserver and one ClamAV-Server. All machines are double-processor intel-xeon 2GHz with 2gb ram and sendmail on it. Currently all incoming mail-servers are sending all mails < 20k to the spam-server via spamass-milter. But spamass-milter is very unstable and I'm planning to use another milter-interface to spamd. MY question: What is the best way to scan mail for viruses and spam in my case ? If I use mimedefang on every mailserver, I have to run spamassassin on every machine, because mimedefang has no support for spamd. If I run mimedefang on the spamserver and use a tcp-socket to connect to clamd I have a single-point-of failure (mimedefang) If mimedefang have a problem, I have no spam and no anti-virus functionality. If I run mimedefang on the spamserver and on the anti-virus-server I have to define two sendmail-milter-sockets, what is no problem, but What is with the X-mimedefang headers ? ... they will be removed and replaced with the second x-mimedefang headers. And I have to run 2 separate configurations. IS there better way to realise something like this ? Thx, florian _______________________________________ Florian Meister Systembetreuer f�r 1036hallo Vorarlberger Medienhaus Gutenbergstrasse 1 6858 Schwarzach Tel. 05572 501 134 Fax 05572 501 97134 E-Mail: [EMAIL PROTECTED] Web: www.medienhaus.at _______________________________________________ Visit http://www.mimedefang.org and http://www.canit.ca MIMEDefang mailing list [EMAIL PROTECTED] http://lists.roaringpenguin.com/mailman/listinfo/mimedefang
