bind9-libs/doc/arm/Bv9ARM.ch07.html

406 lines
20 KiB
HTML
Raw Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<!--
- Copyright (C) 2000-2020 Internet Systems Consortium, Inc. ("ISC")
-
- This Source Code Form is subject to the terms of the Mozilla Public
- License, v. 2.0. If a copy of the MPL was not distributed with this
- file, You can obtain one at http://mozilla.org/MPL/2.0/.
-->
<html lang="en">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=ISO-8859-1">
<title>Chapter 7. BIND 9 Security Considerations</title>
<meta name="generator" content="DocBook XSL Stylesheets V1.78.1">
<link rel="home" href="Bv9ARM.html" title="BIND 9 Administrator Reference Manual">
<link rel="up" href="Bv9ARM.html" title="BIND 9 Administrator Reference Manual">
<link rel="prev" href="Bv9ARM.ch06.html" title="Chapter 6. BIND 9 Configuration Reference">
<link rel="next" href="Bv9ARM.ch08.html" title="Chapter 8. Troubleshooting">
</head>
<body bgcolor="white" text="black" link="#0000FF" vlink="#840084" alink="#0000FF">
<div class="navheader">
<table width="100%" summary="Navigation header">
<tr><th colspan="3" align="center">Chapter 7. <acronym class="acronym">BIND</acronym> 9 Security Considerations</th></tr>
<tr>
<td width="20%" align="left">
<a accesskey="p" href="Bv9ARM.ch06.html">Prev</a> </td>
<th width="60%" align="center"> </th>
<td width="20%" align="right"> <a accesskey="n" href="Bv9ARM.ch08.html">Next</a>
</td>
</tr>
</table>
<hr>
</div>
<div class="chapter">
<div class="titlepage"><div><div><h1 class="title">
<a name="Bv9ARM.ch07"></a>Chapter 7. <acronym class="acronym">BIND</acronym> 9 Security Considerations</h1></div></div></div>
<div class="toc">
<p><b>Table of Contents</b></p>
<dl class="toc">
<dt><span class="section"><a href="Bv9ARM.ch07.html#Access_Control_Lists">Access Control Lists</a></span></dt>
<dt><span class="section"><a href="Bv9ARM.ch07.html#chroot_and_setuid"><span class="command"><strong>Chroot</strong></span> and <span class="command"><strong>Setuid</strong></span></a></span></dt>
<dd><dl>
<dt><span class="section"><a href="Bv9ARM.ch07.html#chroot">The <span class="command"><strong>chroot</strong></span> Environment</a></span></dt>
<dt><span class="section"><a href="Bv9ARM.ch07.html#setuid">Using the <span class="command"><strong>setuid</strong></span> Function</a></span></dt>
</dl></dd>
<dt><span class="section"><a href="Bv9ARM.ch07.html#dynamic_update_security">Dynamic Update Security</a></span></dt>
</dl>
</div>
<div class="section">
<div class="titlepage"><div><div><h2 class="title" style="clear: both">
<a name="Access_Control_Lists"></a>Access Control Lists</h2></div></div></div>
<p>
Access Control Lists (ACLs) are address match lists that
you can set up and nickname for future use in
<span class="command"><strong>allow-notify</strong></span>, <span class="command"><strong>allow-query</strong></span>,
<span class="command"><strong>allow-query-on</strong></span>, <span class="command"><strong>allow-recursion</strong></span>,
<span class="command"><strong>blackhole</strong></span>, <span class="command"><strong>allow-transfer</strong></span>,
<span class="command"><strong>match-clients</strong></span>, etc.
</p>
<p>
Using ACLs allows you to have finer control over who can access
your name server, without cluttering up your config files with huge
lists of IP addresses.
</p>
<p>
It is a <span class="emphasis"><em>good idea</em></span> to use ACLs, and to
control access to your server. Limiting access to your server by
outside parties can help prevent spoofing and denial of service
(DoS) attacks against your server.
</p>
<p>
ACLs match clients on the basis of up to three characteristics:
1) The client's IP address; 2) the TSIG or SIG(0) key that was
used to sign the request, if any; and 3) an address prefix
encoded in an EDNS Client Subnet option, if any.
</p>
<p>
Here is an example of ACLs based on client addresses:
</p>
<pre class="programlisting">
// Set up an ACL named "bogusnets" that will block
// RFC1918 space and some reserved space, which is
// commonly used in spoofing attacks.
acl bogusnets {
0.0.0.0/8; 192.0.2.0/24; 224.0.0.0/3;
10.0.0.0/8; 172.16.0.0/12; 192.168.0.0/16;
};
// Set up an ACL called our-nets. Replace this with the
// real IP numbers.
acl our-nets { x.x.x.x/24; x.x.x.x/21; };
options {
...
...
allow-query { our-nets; };
allow-recursion { our-nets; };
...
blackhole { bogusnets; };
...
};
zone "example.com" {
type master;
file "m/example.com";
allow-query { any; };
};
</pre>
<p>
This allows authoritative queries for "example.com" from any
address, but recursive queries only from the networks specified
in "our-nets", and no queries at all from the networks
specified in "bogusnets".
</p>
<p>
In addition to network addresses and prefixes, which are
matched against the source address of the DNS request, ACLs
may include <code class="option">key</code> elements, which specify the
name of a TSIG or SIG(0) key, or <code class="option">ecs</code>
elements, which specify a network prefix but are only matched
if that prefix matches an EDNS client subnet option included
in the request.
</p>
<p>
The EDNS Client Subnet (ECS) option is used by a recursive
resolver to inform an authoritative name server of the network
address block from which the original query was received, enabling
authoritative servers to give different answers to the same
resolver for different resolver clients. An ACL containing
an element of the form
<span class="command"><strong>ecs <em class="replaceable"><code>prefix</code></em></strong></span>
will match if a request arrives in containing an ECS option
encoding an address within that prefix. If the request has no
ECS option, then "ecs" elements are simply ignored. Addresses
in ACLs that are not prefixed with "ecs" are matched only
against the source address.
</p>
<div class="note" style="margin-left: 0.5in; margin-right: 0.5in;">
<h3 class="title">Note</h3>
<p>
(Note: The authoritative ECS implementation in
<span class="command"><strong>named</strong></span> is based on an early version of the
specification, and is known to have incompatibilities with
other implementations. It is also inefficient, requiring
a separate view for each client subnet to be sent different
answers, and it is unable to correct for overlapping subnets in
the configuration. It can be used for testing purposes, but is
not recommended for production use.)
</p>
</div>
<p>
When <acronym class="acronym">BIND</acronym> 9 is built with GeoIP support,
ACLs can also be used for geographic access restrictions.
This is done by specifying an ACL element of the form:
<span class="command"><strong>geoip [<span class="optional">db <em class="replaceable"><code>database</code></em></span>] <em class="replaceable"><code>field</code></em> <em class="replaceable"><code>value</code></em></strong></span>
</p>
<p>
The <em class="replaceable"><code>field</code></em> indicates which field
to search for a match. Available fields are "country",
"region", "city", "continent", "postal" (postal code),
"metro" (metro code), "area" (area code), "tz" (timezone),
"isp", "asnum", and "domain".
</p>
<p>
<em class="replaceable"><code>value</code></em> is the value to search
for within the database. A string may be quoted if it
contains spaces or other special characters. An "asnum"
search for autonomous system number can be specified using
the string "ASNNNN" or the integer NNNN.
When "country" search is specified with a string is two
characters long, then it must be a standard ISO-3166-1
two-letter country code; otherwise it is interpreted as
the full name of the country. Similarly, if this is a
"region" search and the string is two characters long,
then it treated as a standard two-letter state or province
abbreviation; otherwise it treated as the full name of the
state or province.
</p>
<p>
The <em class="replaceable"><code>database</code></em> field indicates which
GeoIP database to search for a match. In most cases this is
unnecessary, because most search fields can only be found in
a single database. However, searches for "continent" or "country"
can be answered from either the "city" or "country" databases,
so for these search types, specifying a
<em class="replaceable"><code>database</code></em>
will force the query to be answered from that database and no
other. If <em class="replaceable"><code>database</code></em> is not
specified, then these queries will be answered from the "city",
database if it is installed, or the "country" database if it
is installed, in that order. Valid database names are
"country", "city", "asnum", "isp", and "domain". (If using
the legacy GeoIP API, "netspeed" and "org" databases are also
available.)
</p>
<p>
By default, if a DNS query includes an EDNS Client Subnet (ECS)
option which encodes a non-zero address prefix, then GeoIP ACLs
will be matched against that address prefix. Otherwise, they
are matched against the source address of the query. To
prevent GeoIP ACLs from matching against ECS options, set
the <span class="command"><strong>geoip-use-ecs</strong></span> to <code class="literal">no</code>.
</p>
<p>
Some example GeoIP ACLs:
</p>
<pre class="programlisting">geoip country US;
geoip country JP;
geoip db country country Canada;
geoip region WA;
geoip city "San Francisco";
geoip region Oklahoma;
geoip postal 95062;
geoip tz "America/Los_Angeles";
geoip org "Internet Systems Consortium";
</pre>
<p>
ACLs use a "first-match" logic rather than "best-match":
if an address prefix matches an ACL element, then that ACL
is considered to have matched even if a later element would
have matched more specifically. For example, the ACL
<span class="command"><strong> { 10/8; !10.0.0.1; }</strong></span> would actually
match a query from 10.0.0.1, because the first element
indicated that the query should be accepted, and the second
element is ignored.
</p>
<p>
When using "nested" ACLs (that is, ACLs included or referenced
within other ACLs), a negative match of a nested ACL will
the containing ACL to continue looking for matches. This
enables complex ACLs to be constructed, in which multiple
client characteristics can be checked at the same time. For
example, to construct an ACL which allows queries only when
it originates from a particular network <span class="emphasis"><em>and</em></span>
only when it is signed with a particular key, use:
</p>
<pre class="programlisting">
allow-query { !{ !10/8; any; }; key example; };
</pre>
<p>
Within the nested ACL, any address that is
<span class="emphasis"><em>not</em></span> in the 10/8 network prefix will
be rejected, and this will terminate processing of the
ACL. Any address that <span class="emphasis"><em>is</em></span> in the 10/8
network prefix will be accepted, but this causes a negative
match of the nested ACL, so the containing ACL continues
processing. The query will then be accepted if it is signed
by the key "example", and rejected otherwise. The ACL, then,
will only matches when <span class="emphasis"><em>both</em></span> conditions
are true.
</p>
</div>
<div class="section">
<div class="titlepage"><div><div><h2 class="title" style="clear: both">
<a name="chroot_and_setuid"></a><span class="command"><strong>Chroot</strong></span> and <span class="command"><strong>Setuid</strong></span>
</h2></div></div></div>
<p>
On UNIX servers, it is possible to run <acronym class="acronym">BIND</acronym>
in a <span class="emphasis"><em>chrooted</em></span> environment (using
the <span class="command"><strong>chroot()</strong></span> function) by specifying
the <code class="option">-t</code> option for <span class="command"><strong>named</strong></span>.
This can help improve system security by placing
<acronym class="acronym">BIND</acronym> in a "sandbox", which will limit
the damage done if a server is compromised.
</p>
<p>
Another useful feature in the UNIX version of <acronym class="acronym">BIND</acronym> is the
ability to run the daemon as an unprivileged user ( <code class="option">-u</code> <em class="replaceable"><code>user</code></em> ).
We suggest running as an unprivileged user when using the <span class="command"><strong>chroot</strong></span> feature.
</p>
<p>
Here is an example command line to load <acronym class="acronym">BIND</acronym> in a <span class="command"><strong>chroot</strong></span> sandbox,
<span class="command"><strong>/var/named</strong></span>, and to run <span class="command"><strong>named</strong></span> <span class="command"><strong>setuid</strong></span> to
user 202:
</p>
<p>
<strong class="userinput"><code>/usr/local/sbin/named -u 202 -t /var/named</code></strong>
</p>
<div class="section">
<div class="titlepage"><div><div><h3 class="title">
<a name="chroot"></a>The <span class="command"><strong>chroot</strong></span> Environment</h3></div></div></div>
<p>
In order for a <span class="command"><strong>chroot</strong></span> environment
to work properly in a particular directory (for example,
<code class="filename">/var/named</code>), you will need to set
up an environment that includes everything
<acronym class="acronym">BIND</acronym> needs to run. From
<acronym class="acronym">BIND</acronym>'s point of view,
<code class="filename">/var/named</code> is the root of the
filesystem. You will need to adjust the values of
options like <span class="command"><strong>directory</strong></span> and
<span class="command"><strong>pid-file</strong></span> to account for this.
</p>
<p>
Unlike with earlier versions of BIND, you typically will
<span class="emphasis"><em>not</em></span> need to compile <span class="command"><strong>named</strong></span>
statically nor install shared libraries under the new root.
However, depending on your operating system, you may need
to set up things like
<code class="filename">/dev/zero</code>,
<code class="filename">/dev/random</code>,
<code class="filename">/dev/log</code>, and
<code class="filename">/etc/localtime</code>.
</p>
</div>
<div class="section">
<div class="titlepage"><div><div><h3 class="title">
<a name="setuid"></a>Using the <span class="command"><strong>setuid</strong></span> Function</h3></div></div></div>
<p>
Prior to running the <span class="command"><strong>named</strong></span> daemon,
use
the <span class="command"><strong>touch</strong></span> utility (to change file
access and
modification times) or the <span class="command"><strong>chown</strong></span>
utility (to
set the user id and/or group id) on files
to which you want <acronym class="acronym">BIND</acronym>
to write.
</p>
<div class="note" style="margin-left: 0.5in; margin-right: 0.5in;">
<h3 class="title">Note</h3>
<p>
If the <span class="command"><strong>named</strong></span> daemon is running as an
unprivileged user, it will not be able to bind to new restricted
ports if the server is reloaded.
</p>
</div>
</div>
</div>
<div class="section">
<div class="titlepage"><div><div><h2 class="title" style="clear: both">
<a name="dynamic_update_security"></a>Dynamic Update Security</h2></div></div></div>
<p>
Access to the dynamic
update facility should be strictly limited. In earlier versions of
<acronym class="acronym">BIND</acronym>, the only way to do this was
based on the IP
address of the host requesting the update, by listing an IP address
or
network prefix in the <span class="command"><strong>allow-update</strong></span>
zone option.
This method is insecure since the source address of the update UDP
packet
is easily forged. Also note that if the IP addresses allowed by the
<span class="command"><strong>allow-update</strong></span> option include the
address of a slave
server which performs forwarding of dynamic updates, the master can
be
trivially attacked by sending the update to the slave, which will
forward it to the master with its own source IP address causing the
master to approve it without question.
</p>
<p>
For these reasons, we strongly recommend that updates be
cryptographically authenticated by means of transaction signatures
(TSIG). That is, the <span class="command"><strong>allow-update</strong></span>
option should
list only TSIG key names, not IP addresses or network
prefixes. Alternatively, the new <span class="command"><strong>update-policy</strong></span>
option can be used.
</p>
<p>
Some sites choose to keep all dynamically-updated DNS data
in a subdomain and delegate that subdomain to a separate zone. This
way, the top-level zone containing critical data such as the IP
addresses
of public web and mail servers need not allow dynamic update at
all.
</p>
</div>
</div>
<div class="navfooter">
<hr>
<table width="100%" summary="Navigation footer">
<tr>
<td width="40%" align="left">
<a accesskey="p" href="Bv9ARM.ch06.html">Prev</a> </td>
<td width="20%" align="center"> </td>
<td width="40%" align="right"> <a accesskey="n" href="Bv9ARM.ch08.html">Next</a>
</td>
</tr>
<tr>
<td width="40%" align="left" valign="top">Chapter 6. <acronym class="acronym">BIND</acronym> 9 Configuration Reference </td>
<td width="20%" align="center"><a accesskey="h" href="Bv9ARM.html">Home</a></td>
<td width="40%" align="right" valign="top"> Chapter 8. Troubleshooting</td>
</tr>
</table>
</div>
<p xmlns:db="http://docbook.org/ns/docbook" style="text-align: center;">BIND 9.11.16 (Extended Support Version)</p>
</body>
</html>