spamassassin whitelist doesn"t work
am 22.02.2006 17:42:58 von bakerI have a Mandrake [Mandriva] Linux 10.2 box with kernel 2.6.11-12mdk
installed, and sendmail works fine with the following installed
to my satisfaction:
sendmail-devel-8.13.3-2mdk
sendmail-8.13.3-2mdk
sendmail-cf-8.13.3-2mdk
sendmail-doc-8.13.3-2mdk
Spamassassin is installed with the following RPMs and _mostly_
works OK but occassionally an e-mail from a whitelisted address
is filtered away:
spamassassin-3.0.4-0.2.102mdk
spamassassin-tools-3.0.4-0.2.102mdk
spamassassin-spamd-3.0.4-0.2102mdk
spamassassin-spamc-3.0.4-0.2.102mdk
/etc/mail/spamassassin/local.cf looks like this for the
site-specific entries:
def_whitelist_from_rcvd *@*.mit.edu
def_whitelist_from_rcvd *@athena.mit.edu
def_whitelist_from_rcvd *@ATHENA.MIT.edu
def_whitelist_from_rcvd *@mit.edu mit.edu
def_whitelist_from_rcvd *@MIT.EDU
whitelist_from *@mit.edu
whitelist_from *@MIT.EDU
whitelist_from *@mit.EDU
whitelist_from *@*.mit.edu
whitelist_from *@*.MIT.edu
whitelist_from *@*.MIT.EDU
whitelist_from *@*.mit.EDU
As you can see I have accounted for several permutations in the above;
don't know if that's necessary or not.
If an email from a user@MIT.edu is filtered away (several have
turned up in the filtered messages file, but probably 98% make it
through), why would that happen? Shouldn't the above entries ensure
that it will all make it through without being filtered away?
Also I am aware of the .spamassassin directory, but no user has one.
Is it supposed to be automatically generated? If using NIS and
NFS-mounted home directories, will a whitelist file in the .spamassassin
directory still be seen? And presently only root runs the spamassassin
processes, namely:
[root@hostname_deleted ~]# ps -aef| grep spam
root 6775 1 0 Feb18 ? 00:00:00 /usr/bin/spamd -d -c -m5 -H
root 9787 6775 0 Feb21 ? 00:00:09 spamd child
root 22110 6775 0 00:48 ? 00:00:00 spamd child
Should individual users have their own running spamassassin processes
or are the above processes the only ones I need?
Please help; thanks in advance to anyone who does.
baker@deslab.MIT.edu