spam control

[rmedidi]

Hi every body...can you give me all possible ways of getting spam into our netwok. Also please try to give the ways to block it by methods other than Spamassistant.

Thanx
best regards RavI

[Steven E. Protter]

Shalom rmedidi,

chkconfig spamassassin on
service spamassassin on

There are a number of ways of hardening sendmail spam protection at http://www.sendmail.org

SEP

[Ivan Ferreira]

Here is a good link about spam:

http://www-128.ibm.com/developerworks/linux/library/l-spamf.html

[Bill Thorsteinson]

I find greylisting blocks 90% of spam.
Greylisting works by putting the triplet
of sending IP, sending user, and destination
user on a list. Email won't be accepted
from the triplet until a timeout has expired
(say for the first hour). After that email
will be accepted if it is retried withing
a second timeout (say 8 hours). Once email
is accepted the address remains on a white
list for a certain number of days after the
last triplet was last seen (say two months).

Proper email servers will retry and the mail should be delivered with the two timeouts.
Once the triplet is whitelisted there is
no delay.

If the spam is coming from a spambot, then
the spambot will fail on the first attempt.
As they never retry you will have knocked
off a large (I get 90%+) of your spam.

The greylist delay also gives the various
spam traps time to identify the spammers
so you can identify them quicker.

There are a number of services which run
spam traps publish lists of IP addresses
from which spam is originating. Policies
vary so you will want to choose carefully.

Another check is sender policy framework.
This is based on a DNS txt record which
identifies valid sending servers for a
domain. Any e-mail which fails this is
most likely spam. However, some spammers
follow correct SPF policy. Unfortuneately
for them this makes them much easier to
track down.

[Pau Garcia i Quiles]

Try CRM114: http://crm114.sourceforge.net/

It's really good (very accurate).