[Mailman-Users] AMaVis and Mailman

Evilio del Rio edelrio at icm.csic.es
Wed May 17 14:16:31 CEST 2000


Hello,


We are planning to install a mail server (sendmail) with the mailing lists
runned by mailman (http://www.list.org/). We have implemented also an
antivirus scanner for the attachments, AMaVis (http://www.amavis.org/).

Now the problem arises when some user sends a message to one of our
lists with an infected attachment. We have mainly internal local lists
with ~200 users.

When the message arrives it is accepted inmediately by MailMan which, in
turns, tries to distribute the message to the recipients on the list. At
this moment, the mailqueue has only one (big) entry, like this:


QAA23445     1653 Tue May 16 16:49 mylist-admin at mymachine.mydoma.in
                                   user1 at mymachine.mydoma.in
...(39 more entries)
                                   user41 at mymachine.mydoma.in

Now, sendmail delivers locally to each address with an independent virus
scanner process for each local address. If it contains an infected
attachment, the virus scanner triggers 40 (or whatever) alerts, this means
120 messages (one for sender, recipient and administrator each).

I would like to change this behavior so that the message is checked and
the alert triggered only once. I should say that the best moment should be
when the message is "accepted for delivery", just BEFORE mailman has a
chance to process it.

The other possibilities: 

	a) to make only one scan for each local message
(independent of the number of recipients).

	b) to wrap the program Mailer 
also with the virus scanner (scanmails) in the same way that AMAaVis does
for the local mailer. The standard sendmail.cf entry is:

Mprog,	P=/usr/sbin/smrsh, F=lsDFMoqu9, S=10/30, R=20/40, D=$z:/,
	T=X-Unix,
	A=sh -c $u

and it should read something like this:

Mprog,	P=/usr/sbin/scanmails, lsDFMoqu9, S=10/30, R=20/40, D=$z:/,
	T=X-Unix,
	A=scanmails smrsh sh -c $u


I would appreciate any help since I do not know which strategy should I
use (I am no an expert).

Thanks a lot,

________________________________________________________________
Evilio Jose del Rio Silvan          Institut de Ciencies del Mar
edelrio at icm.csic.es          http://members.es.tripod.de/Evilio/
SETI at Home! http://setiathome.ssl.berkeley.edu/






More information about the Mailman-Users mailing list