blob: bb07035e57b1a1b674f797aae1f4111b8406ed41 (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
|
DESCRIPTION
===========
Rspamd is a complex spam filter that allows to estimate messages by many rules,
statistical data and custom services like URL black lists. Each message is
estimated by rspamd and got so called 'spam score'. According to spam score and
user's settings rspamd send recommended action for this message to MTA. Rspamd
has own unique features among spam filters:
* event driven architecture allowing to process many messages at a time
* flexible syntax of rules allowing to write rules in lua language
* a lot of plugins and rules shipped with rspamd distribution
* highly optimized mail processing advanced statistic
All these features allow rspamd to process messages fast and make good results in
spam filtering.
INSTALLATION
============
Rspamd runs on a Unix like operational systems. FreeBSD users can use ports
collection (mail/rspamd) for rspamd installation. Users of other OSes should
use sources to build and install rspamd. You can obtain rspamd sources at
bitbucket download page: https:bitbucket.org/vstakhov/rspamd/downloads.
Build requirements
==================
Rspamd requires several 3-rd party software to build and run:
* libevent - asynchronous event library
* glib - common purposes library
* gmime - mime parser
* lua - extendable scripting language
* cmake - advanced software build system
You can either install them from sources or (recommended) install using package manager of your system.
Build process
=============
Building of rspamd is quite simple:
$ cmake .
$ make
# make install
After installation binaries, rules, plugins and a sample configuration will be
installed in the target directories (prefixed by /usr/local by default). To
start working with rspamd you should do several steps:
1. Copy a sample configuration $PREFIX/etc/rspamd.xml.sample to
$PREFIX/etc/rspamd.xml
2. Edit rspamd.xml according to your system (described
later).
3. Make a directory for rspamd pid file and data (/var/run/rspamd by
default) and make rspamd user (nobody by default) as owner of rspamd data
directory.
4. Make a directory for rspamd logs (or setup syslog to accept
rspamd log messages)
5. Install start script to a proper place (this step is done
when installing from FreeBSD ports)
6. Start rspamd using start script
If start script is not suitable for your system (now rspamd shipped with start
script for FreeBSD, Debian and RedHat like operational systems) you should
write a start script based on your system's documentation.
FURTHER ACTIONS
===============
You can improve quality of rspamd by learning its statistic module. The easiest
way to do it is to use rspamc client (you can setup a custom email alias to
pipe messages to rspamc)
$ rspamc -P 'q1' -c bayes learn_spam [ file1 [file2 [...]]]
$ rspamc -P 'q1' -c bayes learn_ham [ file1 [file2 [...]]]
Note: you should consider to change default controller's password 'q1' to yours
one specified in controller section of configuration.
Also a system administrator may want to customize rule's weights or actions
thresholds. This can be easily done by editing metric section in the
configuration file.
For writing new rules you can examine the main rspamd documentation and lua api
gude and reference.
REFERENCES
==========
Home site: https://bitbucket.org/vstakhov/rspamd
Downloads: https://bitbucket.org/vstakhov/rspamd/downloads
Wiki: https://bitbucket.org/vstakhov/rspamd/wiki/
|
