doc/en_US.ISO8859-1/articles/ipsec-must/article.sgml
Hiroki Sato 5fa082b1f8 Simplify parameter entities in doctype declaration.
Currently we have articles.ent and books.ent, and
for example, articles.ent can be used by putting the
following lines in the doctype declaration:

 <!ENTITY % articles.ent PUBLIC
            "-//FreeBSD//ENTITIES DocBook FreeBSD Articles Entity Set//EN">
 %articles.ent;

This pulls all of the necessary entities via share/sgml/articles.ent.

The translation teams can customize these entities by redefining
the articles.ent file in <langcode>/share/sgml.  See
ja_JP.eucJP/share/sgml for example.
2004-08-08 13:44:01 +00:00

344 lines
9.4 KiB
Text

<!--
The FreeBSD Documentation Project
$FreeBSD$
-->
<!DOCTYPE article PUBLIC "-//FreeBSD//DTD DocBook V4.1-Based Extension//EN" [
<!ENTITY % articles.ent PUBLIC "-//FreeBSD//ENTITIES DocBook FreeBSD Articles Entity Set//EN">
%articles.ent;
]>
<article>
<articleinfo>
<title>Independent Verification of IPsec Functionality in FreeBSD</title>
<author>
<firstname>David</firstname>
<surname>Honig</surname>
<affiliation>
<address><email>honig@sprynet.com</email></address>
</affiliation>
</author>
<pubdate>3 May 1999</pubdate>
<legalnotice id="trademarks" role="trademarks">
&tm-attrib.freebsd;
&tm-attrib.opengroup;
&tm-attrib.general;
</legalnotice>
<abstract>
<para>You installed IPsec and it seems to be working. How do you
know? I describe a method for experimentally verifying that IPsec is
working.</para>
</abstract>
</articleinfo>
<sect1 id="problem">
<title>The Problem</title>
<para>First, let's assume you have <link linkend="ipsec-install">
installed <emphasis>IPsec</emphasis></link>. How do you know
it is <link linkend="caveat">working</link>? Sure, your
connection will not work if it is misconfigured, and it will work
when you finally get it right. &man.netstat.1; will list it.
But can you independently confirm it?</para>
</sect1>
<sect1 id="solution">
<title>The Solution</title>
<para>First, some crypto-relevant info theory:</para>
<orderedlist>
<listitem>
<para>encrypted data is uniformly distributed, i.e., has maximal
entropy per symbol;</para>
</listitem>
<listitem>
<para>raw, uncompressed data is typically redundant, i.e., has
sub-maximal entropy.</para>
</listitem>
</orderedlist>
<para>Suppose you could measure the entropy of the data to- and
from- your network interface. Then you could see the difference
between unencrypted data and encrypted data. This would be true
even if some of the data in <quote>encrypted mode</quote> was
not encrypted---as the outermost IP header must be, if the
packet is to be routable.</para>
<sect2 id="MUST">
<title>MUST</title>
<para>Ueli Maurer's <quote>Universal Statistical Test for Random
Bit Generators</quote>(<ulink
url="http://www.geocities.com/SiliconValley/Code/4704/universal.pdf">
<acronym>MUST</acronym></ulink>) quickly measures the entropy
of a sample. It uses a compression-like algorithm. <link
linkend="code">The code is given below</link> for a variant
which measures successive (~quarter megabyte) chunks of a
file.</para>
</sect2>
<sect2 id="tcpdump">
<title>Tcpdump</title>
<para>We also need a way to capture the raw network data. A
program called &man.tcpdump.1; lets you do this, if you have
enabled the <emphasis>Berkeley Packet Filter</emphasis>
interface in your <link linkend="kernel">kernel's config
file</link>.</para>
<para>The command</para>
<screen><userinput><command>tcpdump</command> -c 4000 -s 10000 -w <replaceable>dumpfile.bin</replaceable></userinput></screen>
<para>will capture 4000 raw packets to
<replaceable>dumpfile.bin</replaceable>. Up to 10,000 bytes per
packet will be captured in this example.</para>
</sect2>
</sect1>
<sect1 id="experiment">
<title>The Experiment</title>
<para>Here is the experiment:</para>
<procedure>
<step>
<para>Open a window to an IPsec host and another window to an
insecure host.</para>
</step>
<step>
<para>Now start <link linkend="tcpdump">capturing
packets</link>.</para>
</step>
<step>
<para>In the <quote>secure</quote> window, run the &unix;
command &man.yes.1;, which will stream the <literal>y</literal>
character. After a while, stop this. Switch to the
insecure window, and repeat. After a while, stop.</para>
</step>
<step>
<para>Now run <link linkend="code">MUST</link> on the
captured packets. You should see something like the
following. The important thing to note is that the secure
connection has 93% (6.7) of the expected value (7.18), and
the <quote>normal</quote> connection has 29% (2.1) of the
expected value.</para>
<screen>&prompt.user; <userinput>tcpdump -c 4000 -s 10000 -w <replaceable>ipsecdemo.bin</replaceable></userinput>
&prompt.user; <userinput>uliscan <replaceable>ipsecdemo.bin</replaceable></userinput>
Uliscan 21 Dec 98
L=8 256 258560
Measuring file ipsecdemo.bin
Init done
Expected value for L=8 is 7.1836656
6.9396 --------------------------------------------------------
6.6177 -----------------------------------------------------
6.4100 ---------------------------------------------------
2.1101 -----------------
2.0838 -----------------
2.0983 -----------------</screen>
</step>
</procedure>
</sect1>
<sect1 id="caveat">
<title>Caveat</title>
<para>This experiment shows that IPsec <emphasis>does</emphasis>
seem to be distributing the payload data
<emphasis>uniformly</emphasis>, as encryption should. However,
the experiment described here <emphasis>cannot</emphasis>
detect many possible flaws in a system (none of which do I have
any evidence for). These include poor key generation or
exchange, data or keys being visible to others, use of weak
algorithms, kernel subversion, etc. Study the source; know the
code.</para>
</sect1>
<sect1 id="IPsec">
<title>IPsec---Definition</title>
<para>Internet Protocol security extensions to IPv4; required for
IPv6. A protocol for negotiating encryption and authentication
at the IP (host-to-host) level. SSL secures only one application
socket; <application>SSH</application> secures only a login;
<application>PGP</application> secures only a specified file or
message. IPsec encrypts everything between two hosts.</para>
</sect1>
<sect1 id="ipsec-install">
<title>Installing IPsec</title>
<para>Most of the modern versions of FreeBSD have IPsec support
in their base source. So you will probably will need to include
<option>IPSEC</option> option in your kernel config and, after
kernel rebuild and reinstall, configure IPsec connections using
&man.setkey.8; command.</para>
<para>A comprehensive guide on running IPsec on FreeBSD is
provided in <ulink
url="&url.books.handbook;/ipsec.html">FreeBSD
Handbook</ulink>.</para>
</sect1>
<sect1 id="kernel">
<title>src/sys/i386/conf/KERNELNAME</title>
<para>This needs to be present in the kernel config file in order
to be able to capture network data with &man.tcpdump.1;. Be sure
to run &man.config.8; after adding this, and rebuild and
reinstall.</para>
<programlisting>device bpf</programlisting>
</sect1>
<sect1 id="code">
<title>Maurer's Universal Statistical Test (for block size=8
bits)</title>
<para>You can find the same code at <ulink
url="http://www.geocities.com/SiliconValley/Code/4704/uliscanc.txt">
this link</ulink>.</para>
<programlisting>/*
ULISCAN.c ---blocksize of 8
1 Oct 98
1 Dec 98
21 Dec 98 uliscan.c derived from ueli8.c
This version has // comments removed for Sun cc
This implements Ueli M Maurer's "Universal Statistical Test for Random
Bit Generators" using L=8
Accepts a filename on the command line; writes its results, with other
info, to stdout.
Handles input file exhaustion gracefully.
Ref: J. Cryptology v 5 no 2, 1992 pp 89-105
also on the web somewhere, which is where I found it.
-David Honig
honig@sprynet.com
Usage:
ULISCAN filename
outputs to stdout
*/
#define L 8
#define V (1&lt;&lt;L)
#define Q (10*V)
#define K (100 *Q)
#define MAXSAMP (Q + K)
#include &lt;stdio.h&gt;
#include &lt;math.h&gt;
int main(argc, argv)
int argc;
char **argv;
{
FILE *fptr;
int i,j;
int b, c;
int table[V];
double sum = 0.0;
int iproduct = 1;
int run;
extern double log(/* double x */);
printf("Uliscan 21 Dec 98 \nL=%d %d %d \n", L, V, MAXSAMP);
if (argc &lt; 2) {
printf("Usage: Uliscan filename\n");
exit(-1);
} else {
printf("Measuring file %s\n", argv[1]);
}
fptr = fopen(argv[1],"rb");
if (fptr == NULL) {
printf("Can't find %s\n", argv[1]);
exit(-1);
}
for (i = 0; i &lt; V; i++) {
table[i] = 0;
}
for (i = 0; i &lt; Q; i++) {
b = fgetc(fptr);
table[b] = i;
}
printf("Init done\n");
printf("Expected value for L=8 is 7.1836656\n");
run = 1;
while (run) {
sum = 0.0;
iproduct = 1;
if (run)
for (i = Q; run && i &lt; Q + K; i++) {
j = i;
b = fgetc(fptr);
if (b &lt; 0)
run = 0;
if (run) {
if (table[b] &gt; j)
j += K;
sum += log((double)(j-table[b]));
table[b] = i;
}
}
if (!run)
printf("Premature end of file; read %d blocks.\n", i - Q);
sum = (sum/((double)(i - Q))) / log(2.0);
printf("%4.4f ", sum);
for (i = 0; i &lt; (int)(sum*8.0 + 0.50); i++)
printf("-");
printf("\n");
/* refill initial table */
if (0) {
for (i = 0; i &lt; Q; i++) {
b = fgetc(fptr);
if (b &lt; 0) {
run = 0;
} else {
table[b] = i;
}
}
}
}
}</programlisting>
</sect1>
</article>