/ 
Mail-Decency-0.1.5
River stage zero No dependents
/ Mail::Decency

NAME

Mail::Decency - Anti-Spam fighting framework

DESCRIPTION

Mail::Decency is an interface between postfix (MTA), a bunch of policies (eg DNSBL, SPF, ..), multiple content filters (eg DSPAM, Bogofilter, ClamAV, DKIM validation, ...) and a log parser.

It is based on POE and Moose and runs as a daemon with multiple forked instances.

SYNOPSIS

Setting up a new policy server

use Mail::Decency::Policy;
my $policy = Mail::Decency::Policy->new( {
    config => '/etc/decency/policy.yml'
} );
$policy->run;

Setting up a new content filter

use Mail::Decency::ContentFilter;
my $content_filter = Mail::Decency::ContentFilter->new( {
    config => '/etc/decency/content-filter.yml'
} );
$content_filter->run;

Setting up a new syslog parser

use Mail::Decency::LogParser;
my $syslog_parser = Mail::Decency::LogParser->new( {
    config => '/etc/decency/log-parser.yml'
} );
$syslog_parser->run;

INTRODUCTION

For now, decency is in alpha-testing state. Don't use it in production. Also: API changes could occure.

WHY ANOTHER POLICY SERVER OR CONTENT FILTER ?

Well, that is the first reason why: there is no single (open source) application that combines both that i know of. The spam threat has not been decreased over the time, quite the contrary: it has hugely increased. I think it is of the outermost importance to bring those two points of defense as close together as possible.

The second reason: CPAN. The existing (perl implemented) solutions lack imho the modular design for being as extendable as they could be and are not released on CPAN.

The third reason is distributability (is this an English word?). In high traffic environments a distributed (mail server) structure is not uncommon. The content filter and policy server solution should be designed for this scenario.

The fourth and last reason is more personal. Complex and voluminous configuration lead often to human mistakes (or at least on my part). decency tries to have a simplified and maintainable configuration. Then again, maybe the configuration seams much easier to me, because i wrote it..

HOW DOES IT WORK ?

There are three components involved in the process:

1. Policy server

Sits at the very frontier and fights spam before it is received with various measurements.

2. Content filter

Middleware for applying all kinds of content filtering, such as spam filters and virus filter from third party. Also implements some own filters (DKIM, Archive, ..) without external software.

3. Log parser

Mail server log analyses is most important for running a healthy system, this tries to simply this effort.

THE STRUCTURE

------------           --------------           ------------------
| INTERNET | -[SMTP]-> | MAILSERVER | -[SMTP]-> | CONTENT FILTER |
------------           --------------           ------------------
                           |   ^                       |
                           v   |                   [REINJECT]
                         ----------                    v
                         | POLICY |               --------------
                         ----------               | MAILSERVER |
                                                  --------------

  • A mail is to be received from the internet, the mail server does not yet accept it.

  • The mail server applies his own policies and then asks the policy server whether the mail shall pass.

  • The mail server rejects the mail finally or accepts and receive it and delivers it to the content filter.

  • The content filter might reject the mail, which will force the mail server to bounce it, or passes it again and re-inject it into another mail server process.

  • The mail server delivers the mail (eg mailbox)

SEE ALSO

AUTHOR

Ulrich Kautz <uk@fortrabbit.de>

COPYRIGHT

Copyright (c) 2010 the "AUTHOR" as listed above

LICENCSE

This library is free software and may be distributed under the same terms as perl itself.