94
|
1 <html>
|
|
2
|
|
3 <head>
|
|
4 <meta http-equiv="Content-Type" content="text/html; charset=windows-1252">
|
|
5 <title>DNSBL Sendmail milter - Version 5.5</title>
|
|
6 </head>
|
|
7
|
|
8 <center>Introduction</center>
|
|
9 <p>This milter is released under the GPL license version 2 included in
|
|
10 the LICENSE file in the distribution, and also available at
|
|
11 <a href="http://www.gnu.org/licenses/gpl.html">http://www.gnu.org/licenses/gpl.html</a>
|
|
12
|
|
13 <p>Consider the case of a mail server that is acting as secondary MX for
|
|
14 a collection of clients, each of which has a collection of mail domains.
|
|
15 Each client may use their own collection of DNSBLs on their primary mail
|
|
16 server. We present here a mechanism whereby the backup mail server can
|
|
17 use the correct set of DNSBLs for each recipient for each message. As a
|
|
18 side-effect, it gives us the ability to customize the set of DNSBLs on a
|
|
19 per-recipient basis, so that fred@example.com could use SPEWS and the
|
|
20 SBL, where all other users @example.com use only the SBL.
|
|
21
|
|
22 <p>This milter can also verify the envelope from/recipient pairs with
|
|
23 the primary MX server. This allows the backup mail servers to properly
|
|
24 reject mail sent to invalid addresses. Otherwise, the backup mail
|
|
25 servers will accept that mail, and then generate a bounce message when
|
|
26 the message is forwarded to the primary server (and rejected there with
|
|
27 no such user).
|
|
28
|
|
29 <p>This milter will also decode (uuencode, base64, mime, html entity,
|
|
30 url encodings) and scan for HTTP and HTTPS URLs and bare hostnames in
|
|
31 the body of the mail. If any of those host names have A or NS records
|
|
32 on the SBL (or a single configurable DNSBL), the mail will be rejected
|
|
33 unless previously whitelisted. This milter also counts the number of
|
|
34 invalid HTML tags, and can reject mail if that count exceeds your
|
|
35 specified limit.
|
|
36
|
|
37 <p>The DNSBL milter reads a text configuration file (dnsbl.conf) on
|
|
38 startup, and whenever the config file (or any of the referenced include
|
|
39 files) is changed. The entire configuration file is case insensitive.
|
|
40 If the configuration cannot be loaded due to a syntax error, the milter
|
|
41 will log the error and quit. If the configuration cannot be reloaded
|
|
42 after being modified, the milter will log the error and send an email to
|
|
43 root from dnsbl@$hostname. You probably want to added dnsbl@$hostname
|
|
44 to your /etc/mail/virtusertable since otherwise sendmail will reject
|
|
45 that message.
|
|
46
|
|
47 <hr> <center>DCC Issues</center>
|
|
48 <p>If you are also using the <a
|
|
49 href="http://www.rhyolite.com/anti-spam/dcc/">DCC</a> milter, there are
|
|
50 a few considerations. You may need to whitelist senders from the DCC
|
|
51 bulk detector, or from the DNS based lists. Those are two very
|
|
52 different reasons for whitelisting. The former is done thru the DCC
|
|
53 whiteclnt config file, the later is done thru the DNSBL milter config
|
|
54 file.
|
|
55
|
|
56 <p>You may want to blacklist some specific senders or sending domains.
|
|
57 This could be done thru either the DCC (on a global basis, or for a
|
|
58 specific single recipient). We prefer to do such blacklisting via the
|
|
59 DNSBL milter config, since it can be done for a collection of recipient
|
|
60 mail domains. The DCC approach has the feature that you can capture the
|
|
61 entire message in the DCC log files. The DNSBL milter approach has the
|
|
62 feature that the mail is rejected earlier (at RCPT TO time), and the
|
|
63 sending machine just gets a generic "550 5.7.1 no such user" message.
|
|
64
|
|
65 <p>The DCC whiteclnt file can be included in the DNSBL milter config by
|
|
66 the dcc_to and dcc_from statements. This will import the (env_to,
|
|
67 env_from, and substitute mail_host) entries from the DCC config into the
|
|
68 DNSBL config. This allows using the DCC config as the single point for
|
|
69 white/blacklisting.
|
|
70
|
|
71 <p>Consider the case where you have multiple clients, each with their
|
|
72 own mail servers, and each running their own DCC milters. Each client
|
|
73 is using the DCC facilities for envelope from/to white/blacklisting.
|
|
74 Presumably you can use rsync or scp to fetch copies of your clients DCC
|
|
75 whiteclnt files on a regular basis. Your mail server, acting as a
|
|
76 backup MX for your clients, can use the DNSBL milter, and include those
|
|
77 client DCC config files. The envelope from/to white/blacklisting will
|
|
78 be appropriately tagged and used only for the domains controlled by each
|
|
79 of those clients.
|
|
80
|
|
81 <hr> <center>Definitions</center>
|
|
82
|
|
83 <p>CONTEXT - a collection of parameters that defines the filtering
|
|
84 context to be used for a collection of envelope recipient addresses.
|
|
85 The context includes such things as the list of DNSBLs to be used, and
|
|
86 the various content filtering parameters.
|
|
87
|
|
88 <p>DNSBL - a named DNS based blocking list is defined by a dns suffix
|
|
89 (e.g. sbl-xbl.spamhaus.org) and a message string that is used to
|
|
90 generate the "550 5.7.1" smtp error return code. The names of these
|
|
91 DNSBLs will be used to define the DNSBL-LISTs.
|
|
92
|
|
93 <p>DNSBL-LIST - a named list of DNSBLs that will be used for specific
|
|
94 recipients or recipient domains.
|
|
95
|
|
96 <hr> <center>Filtering Procedure</center>
|
|
97
|
|
98 <p>If the client has authenticated with sendmail, the mail is accepted,
|
|
99 the filtering contexts are not used, the dns lists are not checked, and
|
|
100 the body content is not scanned. Otherwise, we follow these steps for
|
|
101 each recipient.
|
|
102
|
|
103 <ol>
|
|
104
|
|
105 <li>The envelope to email address is used to find an initial filtering
|
|
106 context. We first look for a context that specified the full email
|
|
107 address in the env_to statement. If that is not found, we look for a
|
|
108 context that specified the entire domain name of the envelope recipient
|
|
109 in the env_to statement. If that is not found, we look for a context
|
|
110 that specified the user@ part of the envelope recipient in the env_to
|
|
111 statement. If that is not found, we use the first top level context
|
|
112 defined in the config file.
|
|
113
|
|
114 <br><br><li>The initial filtering context may redirect to a child
|
|
115 context based on the values in the initial context's env_from statement.
|
|
116 We look for [1) the full envelope from email address, 2) the domain name
|
|
117 part of the envelope from address, 3) the user@ part of the envelope
|
|
118 from address] in that context's env_from statement, with values that
|
|
119 point to a child context. If such an entry is found, we switch to that
|
|
120 child filtering context.
|
|
121
|
|
122 <br><br><li>We lookup [1) the full envelope from email address, 2) the
|
|
123 domain name part of the envelope from address, 3) the user@ part of the
|
|
124 envelope from address] in the filtering context env_from statement.
|
|
125 That results in one of (white, black, unknown, inherit).
|
|
126
|
|
127 <br><br><li>If the answer is black, mail to this recipient is rejected
|
|
128 with "no such user", and the dns lists are not checked.
|
|
129
|
|
130 <br><br><li>If the answer is white, mail to this recipient is accepted
|
|
131 and the dns lists are not checked.
|
|
132
|
|
133 <br><br><li>If the answer is unknown, we don't reject yet, but the dns
|
|
134 lists will be checked, and the content may be scanned.
|
|
135
|
|
136 <br><br><li>If the answer is inherit, we repeat the envelope from search
|
|
137 in the parent context.
|
|
138
|
|
139 <br><br><li>The dns lists specified in the filtering context are checked
|
|
140 and the mail is rejected if any list has an A record for the standard
|
|
141 dns based lookup scheme (reversed octets of the client followed by the
|
|
142 dns suffix).
|
|
143
|
|
144 <br><br><li>If the mail has not been accepted or rejected yet, and the
|
|
145 filtering context specifies a verification host, and the envelope to
|
|
146 email address is covered by this filtering context, and the verification
|
|
147 host is not our own hostname, we open an smtp conversation with that
|
|
148 verification host. The current envelope from and recipient to values
|
|
149 are passed to that verification host. If we receive a 5xy response
|
|
150 those commands, we reject the current recipient with "no such user".
|
|
151
|
|
152 <br><br><li>If the mail has not been accepted or rejected yet, and the
|
|
153 filtering context enables content filtering, and this is the first such
|
|
154 recipient in this smtp transaction, we set the content filtering
|
|
155 parameters from this context, and enable content filtering for the body
|
|
156 of this message.
|
|
157
|
|
158 </ol>
|
|
159
|
|
160 <p>If content filtering is enabled for this body, the mail text is
|
|
161 decoded (uuencode, base64, mime, html entity, url encodings), scanned
|
|
162 for HTTP and HTTPS URLs, and the first <configurable> host names
|
|
163 are checked for their presence on the single <configurable> DNSBL.
|
|
164 The only known list that is suitable for this purpose is the SBL. If
|
|
165 any of those host names are on that DNSBL (or have nameservers that are
|
|
166 on that list), and it is not on the <configurable> ignore list,
|
|
167 the mail is rejected. We also scan for excessive bad html tags, and if
|
|
168 a <configurable> limit is exceeded, the mail is rejected.
|
|
169
|
|
170 <hr> <center>Sendmail access vs. DNSBL</center>
|
|
171 <p>With the standard sendmail.mc dnsbl FEATURE, the dnsbl checks may be
|
|
172 suppressed by entries in the /etc/mail/access database. For example,
|
|
173 suppose you control a /18 of address space, and have allocated some /24s
|
|
174 to some clients. You have access entries like
|
|
175
|
|
176 <pre>
|
|
177 192.168.4 OK
|
|
178 192.168.17 OK
|
|
179 </pre>
|
|
180
|
|
181 <p>to allow those clients to smarthost thru your mail server. Now if
|
|
182 one of those clients happens get infected with a virus that turns a
|
|
183 machine into an open proxy, and their 192.168.4.45 lands on the SBL-XBL,
|
|
184 you will still wind up allowing that infected machine to smarthost thru
|
|
185 your mail servers.
|
|
186
|
|
187 <p>With this DNSBL milter, the sendmail access database cannot override
|
|
188 the dnsbl checks, so that machine won't be able to send mail to or thru
|
|
189 your smarthost mail server (unless the virus/proxy can use smtp-auth).
|
|
190
|
|
191 <p>Using the standard sendmail features, you would add access entries to
|
|
192 allow hosts on your local network to relay thru your mail server. Those
|
|
193 OK entries in the sendmail access database will override all the dnsbl
|
|
194 checks. With this DNSBL milter, you will need to have the local users
|
|
195 authenticate with smtp-auth to get the same effect. You might find <a
|
|
196 href="http://www.ists.dartmouth.edu/classroom/sendmail-ssl-how-to.php">
|
|
197 these directions</a> helpful for setting up smtp-auth if you are on RH
|
|
198 Linux.
|
|
199
|
|
200 <hr> <center>Installation and configuration</center>
|
|
201 <p>Usage: Note that this has ONLY been tested on Linux, specifically
|
|
202 RedHat Linux. In particular, this milter makes no attempt to understand
|
|
203 IPv6. Your mileage will vary. You will need at a minimum a C++
|
|
204 compiler with a minimally thread safe STL implementation. The
|
|
205 distribution includes a test.cpp program. If it fails this milter won't
|
|
206 work. If it passes, this milter might work.
|
|
207
|
|
208 Fetch <a href="http://www.five-ten-sg.com/util/dnsbl.tar.gz">dnsbl.tar.gz</a>
|
|
209 and
|
|
210
|
|
211 <pre>
|
|
212 tar xfvz dnsbl.tar.gz
|
|
213 bash install.bash
|
|
214 </pre>
|
|
215
|
|
216 Read and understand the contents of that install.bash script before you
|
|
217 run it. It may not be suitable for your system. Modify your
|
|
218 sendmail.mc by removing all the "FEATURE(dnsbl" lines, add the following
|
|
219 line in your sendmail.mc and rebuild the .cf file
|
|
220
|
|
221 <pre>
|
|
222 INPUT_MAIL_FILTER(`dnsbl', `S=local:/var/run/dnsbl/dnsbl.sock, F=T, T=C:30s;S:5m;R:5m;E:5m')
|
|
223 </pre>
|
|
224
|
|
225 Read the sample <a
|
|
226 href="http://www.five-ten-sg.com/dnsbl.conf">/etc/dnsbl/dnsbl.conf</a>
|
|
227 file and modify it to fit your configuration. You can test your
|
|
228 configuration files, and see a readable internal dump of them on stdout
|
|
229 with
|
|
230
|
|
231 <pre>
|
|
232 cd /etc/dnsbl
|
|
233 /usr/sbin/dnsbl -c
|
|
234 </pre>
|
|
235
|
|
236 You can check a specific envelope from/to pair with
|
|
237
|
|
238 <pre>
|
|
239 cd /etc/dnsbl
|
|
240 from="$1" # or your from address
|
|
241 to="$2" # or your to address
|
|
242 /usr/sbin/dnsbl -e "$from"'|'"$to"
|
|
243 </pre>
|
|
244
|
|
245 <hr> <center>Performance issues</center>
|
|
246
|
|
247 <p>Consider a high volume high performance machine running sendmail.
|
|
248 Each sendmail process can do its own dns resolution. Typically, such
|
|
249 dns resolver libraries are not thread safe, and so must be protected by
|
|
250 some sort of mutex in a threaded environment. When we add a milter to
|
|
251 sendmail, we now have a collection of sendmail processes, and a
|
|
252 collection of milter threads.
|
|
253
|
|
254 <p>We will be doing a lot of dns lookups per mail message, and at least
|
|
255 some of those will take many tens of seconds. If all this dns work is
|
|
256 serialized inside the milter, we have an upper limit of about 25K mail
|
|
257 messages per day. That is clearly not sufficient for many sites.
|
|
258
|
|
259 <p>Since we want to do parallel dns resolution across those milter
|
|
260 threads, we add another collection of dns resolver processes. Each
|
|
261 sendmail process is talking to a milter thread over a socket, and each
|
|
262 milter thread is talking to a dns resolver process over another socket.
|
|
263
|
|
264 <p>Suppose we are processing 20 messages per second, and each message
|
|
265 requires 20 seconds of dns work. Then we will have 400 sendmail
|
|
266 processes, 400 milter threads, and 400 dns resolver processes. Of
|
|
267 course that steady state is very unlikely to happen.
|
|
268
|
|
269 <hr> <center>Rejected Ideas</center>
|
|
270
|
|
271 <p>The following ideas have been considered and rejected.
|
|
272
|
|
273 <p>Add max_recipients for each mail domain to the configuration.
|
|
274 Recipients in excess of that limit will be rejected, and all the
|
|
275 recipients in that domain will be removed if there are some other
|
|
276 whitelisted recipients. Current spammers *very* rarely send more than
|
|
277 ten recipients in a single smtp transaction, so this won't stop
|
|
278 any significant amount of spam.
|
|
279
|
|
280 <p>Add poison addresses to the configuration. If any recipient is
|
|
281 poison, all recipients are rejected even if they would be whitelisted,
|
|
282 and the data is rejected if sent. I have a collection of spam trap
|
|
283 addresses that would be suitable for such use. Based on my log files,
|
|
284 any mail to those spam trap addresses is rejected based on either dnsbl
|
|
285 lookups or the DCC. So this won't result in blocking any additional
|
|
286 spam.
|
|
287
|
|
288 <p>Add an option to only allow one recipient if the return path is
|
|
289 empty. Based on my log files, there is no mail that violates this
|
|
290 check.
|
|
291
|
|
292 <p>Reject the mail if the envelope from domain name contains any MX
|
|
293 records pointing to 127.0.0.0/8. I don't see any significant amount of spam
|
|
294 sent with such domain names.
|
|
295
|
|
296
|
|
297 <pre>
|
|
298 $Id$
|
|
299 </pre>
|
|
300 </body>
|
|
301 </html>
|