macssh/lsh/doc/lsh.html

1353 lines
55 KiB
HTML
Executable File
Raw Blame History

<html lang="en"><head>
<title>lsh</title>
<meta http-equiv="Content-Type" content="text/html">
<meta name=description content="lsh">
<meta name=generator content="makeinfo 4.0">
<link href="http://texinfo.org/" rel=generator-home>
</head><body>
<p><hr>
Node:<a name="Top">Top</a>,
Next:<a rel=next href="#Introduction">Introduction</a>,
Previous:<a rel=previous href="#(dir)">(dir)</a>,
Up:<a rel=up href="#(dir)">(dir)</a>
<br>
<h1></h1>
<p>This document describes <code>lsh</code> and related programs. The <code>lsh</code> suite
of programs is intended as a free replacement for the <code>ssh</code> suite of
programs. In turn, <code>ssh</code> was intended as a secure replacement for
the <code>rsh</code> and <code>rlogin</code> programs for remote login over the
Internet.
<p><code>lsh</code> is a component of the <small>GNU</small> system.
<p>This manual explains how to use and hack <code>lsh</code>; it corresponds to
<code>lsh</code> version 1.1.6.
<ul>
<li><a href="#Introduction">Introduction</a>:
<li><a href="#Installation">Installation</a>:
<li><a href="#Getting%20started">Getting started</a>:
<li><a href="#Invoking%20lsh">Invoking lsh</a>:
<li><a href="#Invoking%20lshd">Invoking lshd</a>:
<li><a href="#Terminology">Terminology</a>:
<li><a href="#Concept%20Index">Concept Index</a>:
<p>--- The Detailed Node Listing ---
<p>Introduction
</p><li><a href="#Threats">Threats</a>:
<li><a href="#Features">Features</a>:
<li><a href="#Related%20techniques">Related techniques</a>:
<p>Related programs and techniques
</p><li><a href="#ssh1">ssh1</a>: SSH version 1
<li><a href="#ssh2">ssh2</a>: SSH version 2
<li><a href="#Kerberos">Kerberos</a>: Kerberos
<li><a href="#ipsec">ipsec</a>: IP Sec
<p>Getting started
</p><li><a href="#lsh%20basics">lsh basics</a>: Connection with lsh
<li><a href="#tcpip%20forwarding">tcpip forwarding</a>: Forwarding <small>TCP/IP</small> ports
<li><a href="#lshd%20basics">lshd basics</a>: Starting the lshd deamon
<li><a href="#public-key">public-key</a>: Using public-keys
<li><a href="#srp">srp</a>: Using SRP authentication
<li><a href="#sexp">sexp</a>: Examining keys and other S-exp files.
<li><a href="#Converting%20keys">Converting keys</a>:
<p>Invoking <code>lsh</code>
</p><li><a href="#Algorithm%20options">Algorithms</a>: Selecting algorithms.
<li><a href="#Hostauth%20options">Hostauth options</a>:
<li><a href="#Userauth%20options">Userauth options</a>:
<li><a href="#Action%20options">Actions</a>: What to do after login.
<li><a href="#Verbosity%20options">Messages</a>: Tuning the amount of messages.
</ul>
<p><hr>
Node:<a name="Introduction">Introduction</a>,
Next:<a rel=next href="#Installation">Installation</a>,
Previous:<a rel=previous href="#Top">Top</a>,
Up:<a rel=up href="#Top">Top</a>
<br>
<h1>Introduction</h1>
<p>What is this thing called computer security anyway? Why would you want
to use a program like <code>lsh</code>?
<p>This chapter explains the threats <code>lsh</code> tries to protect you from,
and some of the threats that remain. It also describes some of the
technologies used in <code>lsh</code>.
<p>From time to time in this manual, I will speak about the <dfn>enemy</dfn>.
This means anybody who is trying to eavesdrop or disturb your private
communication. This usage is technical, and it does not imply that the
enemy is somehow morally inferior to you: The enemy may be some awful
criminals trying to eavesdrop on you, or it may be the police trying to
eavesdrop on the same criminals.
<p>The enemy can be a criminal, or a competitor, or your boss who's trying
to find out how much you tell collegues at competing firms. It may be
your own or somebody else's national security officials. Or your
ex-boyfriend who happens to be too curious.
<p>So what can the enemy do to your communications and your privacy?
Remember that just because you're paranoid that doesn't mean that nobody
is trying to get you<small>...</small>
<ul>
<li><a href="#Threats">Threats</a>:
<li><a href="#Features">Features</a>:
<li><a href="#Related%20techniques">Related techniques</a>:
</ul>
<p><hr>
Node:<a name="Threats">Threats</a>,
Next:<a rel=next href="#Features">Features</a>,
Previous:<a rel=previous href="#Introduction">Introduction</a>,
Up:<a rel=up href="#Introduction">Introduction</a>
<br>
<h2>Threats</h2>
<p>When logging in to some other machine via the Internet, either in the
same building or a few continents away, there are several things that
may be under enemy attack.
<dl>
<dt><dfn>Local attacks</dfn>
<dd>The enemy controls your local environment. He or she may be looking over
your shoulder. Your local machine might be cracked. Or there may be some
device planted inside your keyboard transmitting everything you type to
the attacker. About the same problems occur if the attacker has taken
control over your target machine, i.e. the remote machine you have
logged in to.
<br><dt><dfn>Denial-of-service attacks</dfn>
<dd>The enemy has cut your network cable, effectively stopping your
communication. Even without doing physical damage, the enemy may be able
to flood and overload computers or network equipment. Or disrupt network
traffic by sending fake packets to hangup your <small>TCP/IP</small>
connections.
<br><dt><dfn>Passive eavesdropping</dfn>
<dd>The enemy may be able to listen to your communication somewhere along
its path. With the global Internet, it's difficult to predict who might
be able to listen. Internet traffic between buildings just a few hundred
meters apart have been observed temporarily being routed through half a
dozen countries, perhaps a few thousand kilometers.
<p>And even without routing anomalies, it is possible that the enemy has
been able to take control of some nearby machine, and can listen in from
there. Of course, passive eavesdropping is most dangerous if you
transmit cleartext passwords. This is the main reason not to use vanilla
telnet to login to remote systems. Use a telnet with support for
<small>SSL</small> or Kerberos, or use a program like <code>lsh</code> or
<code>ssh</code>.
<p>A passive eavesdropper is assumed not to do anything nasty with your
packets beyond listening to them.
<br><dt><dfn>Name resolution attacks</dfn>
<dd>The translation from symbolic <small>DNS</small> names to numeric
ip-addresses may be controlled by the attacker. In this case, you may
think that you are connecting to a friendly machine, when in fact you
are connecting somewhere else.
<br><dt><dfn>Fake packets</dfn>
<dd>It is fairly easy to fake the source address of an <small>IP</small>-packet,
although it is more difficult to get hold on the replies to the faked
packets. But even without any replies, this can cause serious
problems.
<br><dt><dfn>Man-in-the-middle attack</dfn>
<dd>In this attack, the enemy sits between you and the target. When
communicating with you, he pretends to be the target. When communicating
with the target, he pretends to be you. He also passes all information
on more or less unmodified, so that he is invisible to you and the
target. To mount this attack, the enemy either needs physical access to
some network equipment on the path between you and the target, or he has
been able to fool you to connect to him rather than to the target, for
example by manipulating the <small>DNS</small>-system.
</dl>
<p><code>lsh</code> makes no attempt to protect you from local attacks. You have
to trust the endpoint machines. It seems really difficult to uphold any
security if the local machine is compromised. This is important to keep
in mind in the "visitor"-scenario, where you visit a friend or perhaps an
Internet caf<61> and want to connect to some of the machines at home or at
work. If the enemy has been able to compromize your friend's or the
caf<EFBFBD>'s equipment, you may well be in trouble.
<p>Protection from denial-of-service attacks is also a very difficult
problem, and <code>lsh</code> makes no attempt to protect you from that.
<p>Instead, the aim of <code>lsh</code>, and most serious tools for cryptographic
protection of communications across the net, is to isolate the
vulnerabilities to the communication endpoints. If you know that the
endpoints are safe, the enemy should not be able to compromize your
privacy or communications. Except for denial-of-service attacks (which
at least can't be performed without you noticing it).
<p>First of all, <code>lsh</code> provides protection against passive
eavesdropping. In addition, if you take the appropriate steps to make
sure that hostkeys are properly authenticated, <code>lsh</code> also protects
against man-in-the-middle attacks and in particular against attacks on
the name resolution. In short, you need only trust the security at the
end points: Even if the enemy controls all other network equipment, name
resolution and routing infrastructure, etc, he can't do anything beyond
the denial-of-service attack.
<p>And at last, remember that there is no such thing as absolute security.
You have to estimate the value of that which you are protecting, and
adjust the security measures so that your enemies will not find it worth
the effort to break them.
<p><hr>
Node:<a name="Features">Features</a>,
Next:<a rel=next href="#Related%20techniques">Related techniques</a>,
Previous:<a rel=previous href="#Threats">Threats</a>,
Up:<a rel=up href="#Introduction">Introduction</a>
<br>
<h2>Features</h2>
<p><code>lsh</code> does not only provide more secure replacements for
<code>telnet</code>, <code>rsh</code> and <code>rlogin</code>, it also provides some other
features to make it convenient to communicate securely. But <code>lsh</code> is
still in an early stage of development, so this section is expected to
grow with time. One goal for <code>lsh</code> is to make it reasonable easy to
extend it, without messing with the core security functionality.
<p><code>lsh</code> can be configured to allow login based on a personal key-pair
consisting of a private and a public key, so that you can execute remote
commands without typing your password every time. You can also use
Thomas Wu's Secure Remote Password Protocol (<small>SRP</small>). Kerberos support is
on the wish list but not yet supported (see <a href="#Kerberos">Kerberos</a>).
<p>The public-key authentication methods should also be extended to support
Simple Public Key Infrastructure (<small>SPKI</small>) certificates, including some
mechanism to delegate restricted logins.
<p>Forwarding of arbitrary <small>TCP/IP</small> connections is provided. This
is useful for tunneling otherwise insecure protocols, like telnet and
pop, through an encrypted <code>lsh</code> connection.
<p>Convenient tunneling of <small>X</small> is one of the most impressive features of the
original <code>ssh</code> programs. <code>lsh</code> doesn't do this yet. Other kind
of tunneling that may turn out to be useful include authentication (i.e.
<code>ssh-agent</code>), general forwarding of <small>UDP</small>, and why not also
general <small>IP</small>-tunneling.
<p><hr>
Node:<a name="Related%20techniques">Related techniques</a>,
Previous:<a rel=previous href="#Features">Features</a>,
Up:<a rel=up href="#Introduction">Introduction</a>
<br>
<h2>Related programs and techniques</h2>
<p>This sections describes some other programs and techniques related to
<code>lsh</code>. The ssh family of programs use mostly the same kind of
security as <code>lsh</code>. Kerberos and <small>IPSEC</small> operate quite
differently, in particular when it comes to protection against
man-in-the-middle attacks.
<ul>
<li><a href="#ssh1">ssh1</a>: SSH version 1
<li><a href="#ssh2">ssh2</a>: SSH version 2
<li><a href="#Kerberos">Kerberos</a>: Kerberos
<li><a href="#ipsec">ipsec</a>: IP Sec
</ul>
<p><hr>
Node:<a name="ssh1">ssh1</a>,
Next:<a rel=next href="#ssh2">ssh2</a>,
Previous:<a rel=previous href="#Related%20techniques">Related techniques</a>,
Up:<a rel=up href="#Related%20techniques">Related techniques</a>
<br>
<h3><code>ssh-1.x</code></h3>
<p>The first of the Secure shell programs was Tatu Yl<59>nen's <code>ssh</code>.
The latest of the version 1 series is <code>ssh-1.27</code> which speaks
version 1.5 of the protocol. The "free" version of <code>ssh-1.27</code>
does not allow commercial use without additional licensing, which makes
<code>ssh-1.27</code> non-free software according to Debian's Free Software
Guidelines and the Open Source Definition.
<p>The version 1 protocol has some minor weaknesses, in particular, all
support for using stream ciphers was disabled by default a few versions
back, for security reasons.
<p>There also exists free implementations of <code>ssh-1</code>, for both Unix
and Windows. <code>ossh</code> and later OpenSSH are derived from earlier
version av Tatu Yl<59>nen's <code>ssh</code>, and are free software.
<p>Until <code>lsh</code> becomes stable and well tested, I would recommend using
some implementation of the <code>ssh-1</code> protocol.
<p><hr>
Node:<a name="ssh2">ssh2</a>,
Next:<a rel=next href="#Kerberos">Kerberos</a>,
Previous:<a rel=previous href="#ssh1">ssh1</a>,
Up:<a rel=up href="#Related%20techniques">Related techniques</a>
<br>
<h3><code>ssh-2.x</code></h3>
<p><code>ssh2</code> implements the next generation of the Secure Shell
protocol, the development of which is supervised by the <small>IETF</small>
secsh Working Group. <code>lsh</code> implements the required subset of
this protocol. It is intended to be compatible with the <code>ssh2</code>
series of programs distributed by F-Secure Corporation.
<p>However, the existing versions of <code>ssh2</code> gets some details of the
protocol wrong (probably because it predates the protocol
specification), so there is some amount of bug-compatibility required.
<p>Interoperability between independently developed implementations is one
necessary condition for the <code>ssh-2</code> protocol to become a Proposed
Standard.
<p>The license for F-Secure's <code>ssh2</code> programs is similar to that
for recent versions of <code>ssh1</code>, but with a narrower definition of
"non-commercial use".
<p>Besides <code>lsh</code> there are few free implementations of the
<code>ssh-2</code> protocols, but since May 2000 it is supported also by
OpenSSH.
<p><hr>
Node:<a name="Kerberos">Kerberos</a>,
Next:<a rel=next href="#ipsec">ipsec</a>,
Previous:<a rel=previous href="#ssh2">ssh2</a>,
Up:<a rel=up href="#Related%20techniques">Related techniques</a>
<br>
<h3>Kerberos</h3>
<p>Kerberos is a key distribution system originally developed in the late
1980:s as a part of Project Athena at <small>MIT</small>. Recent development
have been done at The Royal Institute of Technology, Stockholm
(<small>KTH</small>).
<p>Kerberos uses a central trusted ticket-granting server, and requires
less trust on the local machines in the system. It does not use
public-key technology.
<p>Usually, Kerberos support is compiled into applications such as telnet,
ftp and X-clients. The ssh family of programs, on the other hand, tries
to do all needed magic, for instance to forward <small>X</small> securely, and then
provides general <small>TCP/IP</small> forwarding as a kitchen sink.
<p>I believe Kerberos' and lsh's protection against passive eavesdropping
are mostly equivalent. The difference is in the set of machines and
assumptions you have to trust in order to be safe from a
man-in-the-middle attack.
<p>I think the main advantage of <code>lsh</code> over Kerberos is that it is
easier to install and use for on ordinary mortal user. In order to set
up key exchange between two different Kerberos systems (or <dfn>Kerberos
realms</dfn>), the respective system operators need to exchange keys. In the
case of two random users at two random sites, setting up <code>lsh</code> or
some other program in the ssh family is likely easier than to get the
operators to spend time and attention. So <code>lsh</code> should be easier to
use in an anarchistic grass-roots environment.
<p>Another perspective is to combine ssh features like <small>X</small> and
<small>TCP/IP</small> forwarding with authentication based on Kerberos. Such
an arrangement may provide the best of two worlds for those who happen
to have an account at a suitable ticket-granting server.
<p><hr>
Node:<a name="ipsec">ipsec</a>,
Previous:<a rel=previous href="#Kerberos">Kerberos</a>,
Up:<a rel=up href="#Related%20techniques">Related techniques</a>
<br>
<h3><small>IPSEC</small></h3>
<p><small>IPSEC</small> is a set of protocols for protecting general
<small>IP</small> traffic. It is developed by another <small>IETF</small> working
group, and is also a required part of <small>IP</small> version 6.
<p>Again, the main difference between <small>IPSEC</small> and Kerberos and ssh
is the set of machines that have to be secure and the keys that have to
be exchanged in order to avoid man-in-the-middle attacks.
<p>Current protocols and implementations of <small>IPSEC</small> only provide
authentication of machines; there's nothing analogous to the user
authentication in ssh or Kerberos.
<p>On the other hand, <small>IPSEC</small> provides one distinct advantage over
application level encryption. Because <small>IP</small> and <small>TCP</small>
headers are authenticated, it provides protection against some
denial-of-service attacks. In particular, it makes attacks that cause
hangup of a <small>TCP</small> connection considerably more difficult.
<p>So it makes sense to use both <small>IPSEC</small> and some application
level cryptographic protocol.
<p>Also note that it is possible to use the <dfn>Point-to-Point Protocol</dfn>
(<small>PPP</small>) to tunnel arbitrary <small>IP</small> traffic accross an ssh
connection. This arrangement provides some of the functionality of
<small>IPSEC</small>, and is sometimes referred to as "a poor man's Virtual
Private Network".
<p><hr>
Node:<a name="Installation">Installation</a>,
Next:<a rel=next href="#Getting%20started">Getting started</a>,
Previous:<a rel=previous href="#Introduction">Introduction</a>,
Up:<a rel=up href="#Top">Top</a>
<br>
<h1>Installation</h1>
<p>You install <code>lsh</code> with the usual <code>./configure &amp;&amp; make &amp;&amp;
make install</code>. For a full listing of the options you can give to
<code>configure</code>, use <code>./configure --help</code>. For example, use
<code>--without-pty</code> to disable pty-support.
<p>The most commonly used option is <code>--prefix</code>, which tells
configure where lsh should be installed. Default prefix is
<code>/usr/local</code>. The <code>lshd</code> server is installed in
<code>$prefix/sbin</code>, all other programs and scripts are installed in
<code>$prefix/bin</code>.
<p>The configure script tries to figure out if the linker needs any special
flags specifying where to find dynamically linked libraries at run time
(one case where this matters is if you have a dynamic libz.so installed
in a non -standard place). Usually, you can use
<pre>./configure --with-lib-path=/opt/lib:/other/place
</pre>
<p>to specify extra library directories, and the configure script should do
the right thing. If it doesn't work, or you believe that you know your
system better than <code>./configure</code>, just set LDFLAGS and/or
LD_LIBRARY_PATH to the right values instead.
<p><hr>
Node:<a name="Getting%20started">Getting started</a>,
Next:<a rel=next href="#Invoking%20lsh">Invoking lsh</a>,
Previous:<a rel=previous href="#Installation">Installation</a>,
Up:<a rel=up href="#Top">Top</a>
<br>
<h1>Getting started</h1>
<p>This section tells you how to perform some common tasks using the
<code>lsh</code> suite of programs, without covering all options and
possibilities.
<ul>
<li><a href="#lsh%20basics">lsh basics</a>: Connection with lsh
<li><a href="#tcpip%20forwarding">tcpip forwarding</a>: Forwarding <small>TCP/IP</small> ports
<li><a href="#lshd%20basics">lshd basics</a>: Starting the lshd deamon
<li><a href="#public-key">public-key</a>: Using public-keys
<li><a href="#srp">srp</a>: Using SRP authentication
<li><a href="#sexp">sexp</a>: Examining keys and other S-exp files.
<li><a href="#Converting%20keys">Converting keys</a>:
</ul>
<p><hr>
Node:<a name="lsh%20basics">lsh basics</a>,
Next:<a rel=next href="#tcpip%20forwarding">tcpip forwarding</a>,
Previous:<a rel=previous href="#Getting%20started">Getting started</a>,
Up:<a rel=up href="#Getting%20started">Getting started</a>
<br>
<h2><code>lsh</code> basics</h2>
<p><code>lsh</code> is the program you use for connection to a remote machine. A
few examples are:
<pre>lsh sara.lysator.liu.se
</pre>
<p>Connects to <code>sara.lysator.liu.se</code> and starts an interactive shell.
In this example, and in the rest of the examples in this section, lsh
will ask for your password, unless you have public-key user
authentication set up.
<p>The first time you try to connect between two machines, <code>lsh</code>
typically complains about an "unknown host key". This is because it
has no reason to believe that it was the right machine that answered,
and not a machine controlled by the enemy (see <a href="#Threats">Threats</a>). The default
behaviour is to never ever accept a server that is not properly
authenticated. A machine is considered authentic if it follows the
protocol and has its public hostkey listed in <code>~/.lsh/known_hosts</code>.
<p>To make lsh less paranoid, use
<pre>lsh --sloppy-host-authentication sara.lysator.liu.se
</pre>
<p>Then <code>lsh</code> will display a <dfn>fingerprint</dfn> of the host key of the
remote machine, and ask you if it is correct. If so, the machine is
considered authentic and its key is appended to the file
<code>~/.lsh/captured_keys</code>. You can copy keys you have verified to
<code>~/.lsh/known_hosts</code>.
<p>You can even use
<pre>lsh --sloppy-host-authentication --capture-to ~/.lsh/known_hosts
</pre>
<p>to get <code>lsh</code> to behave more like the traditional <code>ssh</code> program.
<p>You can create fingerprints for the hostkeys you need regularly, and
keep with you (see <a href="#sexp">sexp</a>).
<pre>lsh -l omar sara.lysator.liu.se
</pre>
<p>Connects, like above, but tries to log in as the user "omar".
<pre>lsh sara.lysator.liu.se tar cf - some/dir | (cd /target/dir &amp;&amp; tar -xf -)
</pre>
<p>Copies a directory from the remote machine, by executing one remote and
one local <code>tar</code> process and piping them together.
<pre>CVS_RSH=lsh cvs -d sara.lysator.liu.se:/lysator/cvsroot co lsh
</pre>
<p>Checks out the <code>lsh</code> source code from the <small>CVS</small>
repository.
<p><hr>
Node:<a name="tcpip%20forwarding">tcpip forwarding</a>,
Next:<a rel=next href="#lshd%20basics">lshd basics</a>,
Previous:<a rel=previous href="#lsh%20basics">lsh basics</a>,
Up:<a rel=up href="#Getting%20started">Getting started</a>
<br>
<h2>Port forwarding</h2>
<p>One useful feature of <code>lsh</code> and other ssh-like programs is the
ability to forward arbitrary connections inside the encrypted
connection. There are two flavors: "local" and "remote" forwarding.
<p>An example of local forwarding is
<pre>lsh -L 4000:kom.lysator.liu.se:4894 sara.lysator.liu.se
</pre>
<p>This makes <code>lsh</code> listen on port 4000 on the <em>local</em> machine. When
someone connects, <code>lsh</code> asks the server to open a connection from
the <em>remote</em> machine (i.e. <code>sara</code>) to port 4894 on another machine
(i.e. <code>kom</code>). The two connections are piped together using an
encrypted channel.
<p>There are a few things that should be noted here:
<ul>
<li>By default, <code>lsh</code> only listens on the loopback interface, so only
clients on the same machine can use the tunnel. To listen on all
interfaces, use the <code>-g</code> flag.
<li>A connection through the tunnel consists of three parts:
<ol type=1 start=1>
<li>From a client socket to the local port (4000 in this example) that
<code>lsh</code> listens on.
<li>The tunnel itself, from the local machine to the tunnel endpoint,
which is <code>sara</code> in this example.
<li>The connection from the tunnel endpoint to the ultimate target, in this
example from <code>sara</code> to <code>kom</code>.
</ol>
<p>Only the middle part is protected by <code>lsh</code>: all data flowing
through the tunnel is sent across the first and last part <em>in the
clear</em>. So forwarding doesn't offer much protection unless the tunnel
endpoint and the ultimate target machine are close to eachother. They
should usually be either the same machine, or two machines connected by
a local network that is trusted.
</p><li>Port forwarding is very useful for traversing firewalls. Of course, you
don't need to use lsh-style forwarding just to get out, there are other
tools like HTTPTunnel for that. But <code>lsh</code> helps you get out through
the firewall in a secure way.
<li>Port forwarding is done in addition to anything else <code>lsh</code> is
doing. In the example above, a tunnel is set up, but <code>lsh</code> will
also start an interactive shell for you. Just like the <code>-L</code>
option was not present. If this is not what you want, the <code>-N</code>
option is for you (see <a href="#Invoking%20lsh">Invoking lsh</a>)
</ul>
<p>Remote forwarding is similar, but asks the <em>remote</em> machine to
listen on a port. An example of remote forwarding is
<pre>lsh -g -R 8080:localhost:80 sara.lysator.liu.se
</pre>
<p>This asks the remote machine to listen on port 8080 (note that you are
probably not authorized to listen on port 80). Whenever someone
connects, the connection is tunnelled to your local machine, and
directed to port 80 on the same machine. Note the use of <code>-g</code>;
the effect is to allow anybody in the world to use the tunnel to connect
to your local webserver.
<p>The same considerations that apply to forwarded local ports apply also to
forwarded remote ports.
<p>At last, you can use any number of <code>-L</code> and <code>-R</code> options
on the same command line.
<p><hr>
Node:<a name="lshd%20basics">lshd basics</a>,
Next:<a rel=next href="#public-key">public-key</a>,
Previous:<a rel=previous href="#tcpip%20forwarding">tcpip forwarding</a>,
Up:<a rel=up href="#Getting%20started">Getting started</a>
<br>
<h2><code>lshd</code> basics</h2>
<p>There are no global configuration files for <code>lshd</code>; all
configuration is done with command line options (see <a href="#Invoking%20lshd">Invoking lshd</a>).
<p>To run <code>lshd</code>, you must first create a hostkey, usually stored in
<code>/etc/lsh_host_key</code>. To do this, run
<pre>lsh-keygen | lsh-writekey -o /etc/lsh_host_key
</pre>
<p>This will also create a file <code>/etc/lsh_host_key.pub</code>,
containing the corresponding public key.
<p>A typical command line for starting lshd in daemon mode is simply
<pre>lshd --daemonic
</pre>
<p>You can find init script for <code>lshd</code> tailored for Debian's and
RedHat's GNU/Linux systems in the <code>contrib</code> directory.
<p>It is also possible to let <code>init</code> start <code>lshd</code>, by adding it in
<code>/etc/inittab</code>.
<p><hr>
Node:<a name="public-key">public-key</a>,
Next:<a rel=next href="#srp">srp</a>,
Previous:<a rel=previous href="#lshd%20basics">lshd basics</a>,
Up:<a rel=up href="#Getting%20started">Getting started</a>
<br>
<h2>Using public-key user authentication</h2>
<p>Public-key user authentication is a way to authenticate for login,
without having to type any passwords. There are two steps: Creating a
key pair, and authorizing the public key to the systems where you want
to log in.
<p>To create a keypair, run
<pre>lsh-keygen | lsh-writekey
</pre>
<p>This can take some time, but in the end it creates two files
<code>~/.lsh/identity</code> and <code>~/.lsh/identity.pub</code>.
<p>If you want to use the key to login to some other machine, say
<code>sara</code>, you can do that by first copying the key,
<pre>lsh sara.lysator.liu.se '&gt;my-key.pub' &lt;~/.lsh/identity.pub
</pre>
<p>then authorizing it by executing, on <code>sara</code>,
<pre>lsh-authorize my-key.pub
</pre>
<p>For security reasons, you should keep the private key
<code>~/.lsh/identity</code> secret. Anybody who can read that file will be
able to login in your name to any machine where the corresponding public
key is registered as an authorized key.
<p>Naturally, you should also make sure not to authorize any keys but your
own. For instance, it is inappropriate to use an insecure mechanism such
as unauthenticated email, <code>ftp</code> or <code>http</code> to transfer your
public key to the machines where you want to authorize it.
<p>If you have accounts on several systems, you usually create a single
keypair on each of the systems, and on each system you authorize some or
all of your other public keys for login.
<p>Note that <code>lsh</code> does <em>not</em> currently encrypt your private
key in any way. That means that you can lose it if a backup tape gets
into the wrong hands, and if you use NFS it will likely be sent in the
clear across your local network. (Actually, you can encrypt the key
using a pass phrase, by passing the <code>-c 3des</code> option to
<code>lsh-writekey</code>, but <code>lsh</code> doesn't yet support decrypting
it).
<p><hr>
Node:<a name="srp">srp</a>,
Next:<a rel=next href="#sexp">sexp</a>,
Previous:<a rel=previous href="#public-key">public-key</a>,
Up:<a rel=up href="#Getting%20started">Getting started</a>
<br>
<h2>Using <small>SRP</small> authentication</h2>
<p>The Secure Remote Password protocol is a fairly new protocol that
provides mutual authentication based on a password. To use it, you must
first choose a secret password. Next, you create a <dfn>password
verifier</dfn> that is derived from the password. The verifier is stored on
the target machine (i.e. the machine you want to log in to).
<p>To create a verifier, you run the <code>srp-gen</code> program and type
your new password. You have to do it on either the target machine,
redirecting the output to ~/.lsh/srp-verifier, or you can generate it on
some other machine and copy it to the target.
<p>The main advantage of using <small>SRP</small> is that you use the password
not only to get access to the remote machine, but you also use it to
authenticate the remote machine. I.e. you can use it to connect
securely, <em>without</em> having to know any hostkeys or fingerprints
beforehand!
<p>For instance, you could connect using <small>SRP</small> to fetch the hostkey
fingerprint for the remote machine, as a kind of bootstrapping
procedure, and then use traditional authentication methods for further
connections.
<p>For this to work, the verifier <em>must</em> be kept <em>secret</em>. If the
enemy gets your verifier, he can mount some attacks:
<ul>
<li>He can mount a <dfn>dictionary attack</dfn> on your password, i.e. generate a large
list of likely password and check if any of them matches yours.
<li>He can impersonate the server. That means that if you try to connect to
the remote machine using <small>SRP</small>, and the attacker can intercept
your connection (e.g. by attacking the name resolution or routing
system) he can successfully pretend to be the real server.
</ul>
<p>If you use <small>SRP</small> to get the hostkey or fingerprint for the
remote machine, as outlined above, the impersonation attack destroys
security, you could just as well connect the hostkey presented by the
remote server without verifying it at all.
<p>If you use <small>SRP</small> exclusively, the situation seems somewhat
different. As far as I can see, an attacker knowing your verifier can
not mount a traditional man-in-the-middle-attack: He can play the
server's part when talking to you, but in order to play your part when
talking to the real server, he needs to know your password as well.
<p><small>SRP</small> support is disabled by default, but can be enabled by the
<code>--srp-keyexchange</code> option to <code>lshd</code> and <code>lsh</code>
(naturally, it won't be used unless enabled on both sides). At the time
of this writing, <small>SRP</small> is too new to be trusted, at least
according to conservative cryptographers (and remember that conservatism
is a virtue when it comes to security).
<p>And even if <small>SRP</small> in itself is secure, the way <code>lsh</code>
integrates it into the <code>ssh</code> protocol has not had much peer review.
The bottom line of this disclaimer is that the <small>SRP</small> support in
<code>lsh</code> should be considered experimental.
<p>As far as I know, <small>SRP</small> is not supported by any other <code>ssh</code>
implementation. The protocol <code>lsh</code> uses is described in the
<code>doc/srp-spec.txt</code>.
<p><hr>
Node:<a name="sexp">sexp</a>,
Next:<a rel=next href="#Converting%20keys">Converting keys</a>,
Previous:<a rel=previous href="#srp">srp</a>,
Up:<a rel=up href="#Getting%20started">Getting started</a>
<br>
<h2>Examining keys and other sexp files</h2>
<p>Keys and most other objects <code>lsh</code> needs to store on disk are
represented as so called S-expressions or <dfn>sexps</dfn> for short.
S-expressions have their roots in the Lisp world, and a variant of them
are used in the Simple Public Key Infrastructure (<small>SPKI</small>).
Currently, <code>lsh</code>'s support for <small>SPKI</small> is quite limited,
but it uses <small>SPKI</small>'s formats for keys and Access Control Lists
(<small>ACL</small>:s).
<p>There are several flavours of the sexp syntax:
<ul>
<li>The canonical syntax is somewhere between a text and a binary format,
and is extremely easy for programs to read and write.
<li>The transport syntax, which is suitable when embedding sexps in text
files. It is essentially the canonical representation, encoded using
base64.
<li>The advanced syntax, which is intended for humans to read and write, and
bears some resemblance to Lisp expressions.
</ul>
<p>To see what your <code>~/.lsh/known_hosts</code> file really contains, try
<pre>sexp-conv -i advanced &lt; ~/.lsh/known_hosts
</pre>
<p>The <code>sexp-conv</code> program can also be used to computes
fingerprints. The fingerprint of a key (or any sexp, for that matter) is
simply the hash of its canonical representation. For example,
<pre>sexp-conv --raw-hash &lt;/etc/lsh_host_key.pub
</pre>
<p><hr>
Node:<a name="Converting%20keys">Converting keys</a>,
Previous:<a rel=previous href="#sexp">sexp</a>,
Up:<a rel=up href="#Getting%20started">Getting started</a>
<br>
<h2>Converting keys from <code>ssh2</code> and OpenSSH</h2>
<p>If you are already using <code>ssh2</code> or OpenSSH, and have creating
one or more personal keypairs, you need to convert the public keys to
<code>lsh</code>'s format before you can authorize them. Use the supplied
<code>ssh-conv</code> script,
<pre>ssh-conv &lt;openssh-key.pub &gt;new-key.pub
</pre>
<p>You can then use the usual <code>lsh-authorize</code> on the converted
keys. For this to work, you must be using <small>DSA</small> keys:
<code>lsh</code> doesn't support <code>RSA</code>.
<p>There is currently no tools for converting private keys.
<p><hr>
Node:<a name="Invoking%20lsh">Invoking lsh</a>,
Next:<a rel=next href="#Invoking%20lshd">Invoking lshd</a>,
Previous:<a rel=previous href="#Getting%20started">Getting started</a>,
Up:<a rel=up href="#Top">Top</a>
<br>
<h1>Invoking <code>lsh</code></h1>
<p><a name="lsh-usage"></a>
<p>You use <code>lsh</code> to login to a remote machine. Basic usage is
<p><code>lsh [-p <var>port number</var>] sara.lysator.liu.se</code>
<p>which attempts to connect, login, and start an interactive shell on the
remote machine. Default <var>port number</var> is whatever your system's
<code>/etc/services</code> lists for <code>ssh</code>. Usually, that is port 22.
<p>There is a plethora of options to <code>lsh</code>, to let you configure where
and how to connect, how to authenticate, and what you want to do once
properly logged in to the remote host. Many options have both long and
short forms. This manual does not list all variants; for a full listing
of supported options, use <code>lsh --help</code>.
<ul>
<li><a href="#Algorithm%20options">Algorithms</a>: Selecting algorithms.
<li><a href="#Hostauth%20options">Hostauth options</a>:
<li><a href="#Userauth%20options">Userauth options</a>:
<li><a href="#Action%20options">Actions</a>: What to do after login.
<li><a href="#Verbosity%20options">Messages</a>: Tuning the amount of messages.
</ul>
<p><hr>
Node:<a name="Algorithm%20options">Algorithm options</a>,
Next:<a rel=next href="#Hostauth%20options">Hostauth options</a>,
Previous:<a rel=previous href="#Invoking%20lsh">Invoking lsh</a>,
Up:<a rel=up href="#Invoking%20lsh">Invoking lsh</a>
<br>
<h2>Algorithm options</h2>
<p>Before a packet is sent, each packet can be compressed, encrypted
and authenticated, in that order. When the packet is received, it is
first decrypted, next it is checked that it is authenticated properly,
and finally it is decompressed. The algorithms used for this are
negotiated with the peer at the other end of the connection, as a part
of the initial handshake and key exchange.
<p>Each party provides a list of supported algorithms, and the first
algorithm listed by the client, which is also found on the server's
list, is selected. Note that this implies that order in which algorithms
are listed on the server's list doesn't matter: if several algorithms
are present on both the server's and the client's lists, it's the
client's order that determines which algorithm is selected.
<p>Algorithms of different types, e.g. data compression and message
authentication, are negotiated independently. Furthermore, algorithms
used for transmission from the client to the server are independent of
the algorithms used for transmission from the server to the client.
There are therefore no less than six different lists that could be
configured at each end.
<p>The command line options for lsh and lshd don't let you specify
arbitrary lists. For instance, you can't specify different preferences
for sending and receiving.
<p>There is a set of default algorithm preferences. When you use a command
line option to say that you want to use <var>algorithm</var> for one of the
algorithms, the default list is replaced with a list containing the
single element <var>algorithm</var>. For example, if you use <code>-c
arcfour</code> to say that you want to use <code>arcfour</code> as the encryption
algorithm, the connection will either end up using <code>arcfour</code>, or
algorithm negotiation will fail because the peer doesn't support
<code>arcfour</code>.
<p><table><tr align="left"><td>Option
<td>Algorithm type <td>Default <td>
<br></tr><tr align="left"><td><code>-z</code> <td>Data compression
<td><code>none</code>, <code>zlib</code>
<td>The default preference list supports zlib compression, but
prefers not to use it.
<br></tr><tr align="left"><td><code>-c</code> <td>Encryption
<td><code>3dec-cbc</code>, <code>blowfish-cbc</code>, <code>arcfour</code>
<code>arcfour</code>
<td>The default encryption algorithm is triple-DES in CBC mode. This
seems to be the algorithm of choice among conservative cryptographers.
The default list includes only quite old and well studied algorithms.
There is a special algorithm name <code>all</code> to enable all supported
encryption algorithms (except <code>none</code>).
<br></tr><tr align="left"><td><code>-m</code> <td>Message Authentication
<td><code>hmac-sha1</code>, <code>hmac-md5</code>
<td>Both supported message authentication algorithms are of the
<small>HMAC</small> family.
<br></tr></table>
<p>As a special case, <code>-z</code> with no argument changes the compression
algorithm list to <code>zlib</code>, <code>none</code>, which means that you want to
use <code>zlib</code> if the other end supports it. This is different from
<code>-zzlib</code> which causes the negotiation to fail if the other end
doesn't support <code>zlib</code>. A somewhat unobvious consequence of
<code>-z</code> having an <em>optional</em> argument is that if you provide an
argument, it must follow directly after the option letter, no spaces
allowed.
<p><hr>
Node:<a name="Hostauth%20options">Hostauth options</a>,
Next:<a rel=next href="#Userauth%20options">Userauth options</a>,
Previous:<a rel=previous href="#Algorithm%20options">Algorithm options</a>,
Up:<a rel=up href="#Invoking%20lsh">Invoking lsh</a>
<br>
<h2>Host authentication options</h2>
<p>As described earlier (see <a href="#Threats">Threats</a>), proper authentication of the
remote host is crucial to protect the connection against
man-in-the-middle attacks. By default, <code>lsh</code> verifies the server's
claimed host key against the <dfn>Access Control Lists</dfn> in
<code>~/.lsh/known_hosts</code>. If the remote host cannot be authenticated,
the connection is dropped.
<p>The options that change this behaviour are
<dl>
<dt><code>--host-db</code>
<dd>Specifies the location of the <small>ACL</small> file.
<br><dt><code>--sloppy-host-authentication</code>
<dd>Tell <code>lsh</code> not to drop the connection if the server's key can not
be authenticated. Instead, it displays the fingerprint of the key, and
asks if it is trusted. The received key is also appended to the file
<code>~/.lsh/captured_keys</code>. If run in quiet mode, <code>lsh -q
--sloppy-host-authentication</code>, <code>lsh</code> connects to any host, no
questions asked.
<br><dt><code>--strict-host-authentication</code>
<dd>Disable sloppy operation (this is the default behaviour).
<br><dt><code>--capture-to</code>
<dd>Use some other file than <code>~/.lsh/captured_keys</code>. For example,
<pre>lsh --sloppy-host-authentication --capture-to ~/.lsh/known_hosts
</pre>
<p>makes <code>lsh</code> behave more like the <code>ssh</code> program.
<br><dt><code>--srp-keyexchange</code>
<dd>Try using <small>SRP</small> for keyexchange and mutual authentication.
</dl>
<p><hr>
Node:<a name="Userauth%20options">Userauth options</a>,
Next:<a rel=next href="#Action%20options">Action options</a>,
Previous:<a rel=previous href="#Hostauth%20options">Hostauth options</a>,
Up:<a rel=up href="#Invoking%20lsh">Invoking lsh</a>
<br>
<h2>User authentication options</h2>
<dl>
<br><dt><code>-l</code>
<dd>Provide a name to use when logging in. By default, the value of the
<code>LOGNAME</code> environment variable is used.
<br><dt><code>-i</code>
<dd>Try the keys from this file to log in. By default, <code>lsh</code> uses
<code>~/.lsh/identity</code>, if it exists. It ought to be possible to use
several <code>-i</code> options to use more than one file, but that is
currently not implemented.
<br><dt><code>--no-publickey</code>
<dd>Don't attempt to log in using public key authentication.
<br><dt><code>--no-userauth</code>
<dd>Attempt to start a session without performing the user authentication
sub-protocol at all. This is the default if <small>SRP</small> was used for
the initial handshake, and should not make much sense otherwise.
</dl>
<p><hr>
Node:<a name="Action%20options">Action options</a>,
Next:<a rel=next href="#Verbosity%20options">Verbosity options</a>,
Previous:<a rel=previous href="#Userauth%20options">Userauth options</a>,
Up:<a rel=up href="#Invoking%20lsh">Invoking lsh</a>
<br>
<h2>Action options</h2>
<p>There are many things <code>lsh</code> can do once you are logged in. There
are two types of options that control this: <dfn>actions</dfn> and
<dfn>action modifiers</dfn>. For short options, actions use uppercase letters
and modifiers use lowercase.
<p>For each modifier <code>--foo</code> there's also a negated form
<code>--no-foo</code>. Options can also be negated be preceding it with the
special option <code>-n</code>. This is mainly useful for negating short
options. For instance, use <code>-nt</code> to tell <code>lsh</code> not to
request a remote pseudo terminal. Each modifier and its negation can be
used several times on the command line. For each action, the latest
previous modifier of each pair apply.
<p>First, the actions:
<dl>
<br><dt><code>-L</code>
<dd>Requests forwarding of a local port. This option takes mandatory
argument of the form
<var>listen-port</var>:<var>target-host</var>:<var>target-port</var>. This option tells
<code>lsh</code> to listen on <var>listen-port</var> on the local machine. When
someone conects to that port, <code>lsh</code> asks the remote server to open
a connection to <var>target-port</var> on <var>target-host</var>, and if it
succeeds, the two connections are joined together through an the
<code>lsh</code> connection. Both port numbers should be given in decimal.
<br><dt><code>-R</code>
<dd>Requests forwarding of a remote port. It takes one mandatory argument,
just like <code>-L</code>. But in this case <code>lsh</code> asks the
<em>remote</em> server to listen on <var>listen-port</var>. When someone
connects to the remote hosts, the server will inform the local
<code>lsh</code>. The local <code>lsh</code> then connects to <var>target-port</var> on
<var>target-host</var>.
<br><dt><code>-E</code>
<dd>This option takes one mandatory argument, which is a command line to be
executed on the remote machine.
<br><dt><code>-S</code>
<dd>Start an interactive shell on the remote machine.
<br><dt><code>-G</code>
<dd>Open a gateway on the local machine. A gateway is a local socket,
located under /tmp, that can be used for controlling and using the ssh
connection. It is protected using the ordinary file permissions. This
feature is experimental.
<br><dt><code>-N</code>
<dd>This is a no-operation action. It inhibits the default action, which is
to start an interactive shell on the remote machine. It is useful if you
want to set up a few forwarded tunnels, and nothing more.
</dl>
<p>If there are trailing arguments after the name of the remote system,
this is equivalent to a <code>-E</code> option, with a command string
constructed by catenating all the remaining arguments, separated by
spaces. This implies that the arguments are usually expanded first by
the local shell, and then the resulting command string is interpreted
again by the remote system.
<p>If there are no trailing arguments after the name of the remote system,
and the <code>-N</code> option is not given, the default action is to start
a shell on the remote machine. I.e. this is equivalent to the
<code>-S</code> option.
<p>There are a few supported modifiers:
<dl>
<br><dt><code>-t</code>
<dd>Request a pseudo terminal. <code>lsh</code> asks the remote system to allocate
a pseudo terminal. If it succeeds, the local terminal is set to raw
mode. The default behaviour is to request a pty if and only if the
local <code>lsh</code> process has a controlling terminal. This modifier
applies to actions that create remote processes, i.e. <code>-E</code> and
<code>-S</code>, as well as the default actions.
<p>Currently, this option is ignored if there is no local terminal.
<br><dt><code>--stdin</code>
<dd>Redirect the stdin of a remote process from a given, local, file.
Default is to use <code>lsh</code>'s stdin for the first process, and
<code>/dev/null</code> for the rest. This option applies to the <code>-E</code>
and <code>-S</code> options as well as to the default actions. The option
applies to only one process; as soon as it is used it is reset to the
default.
<br><dt><code>--stdout</code>
<dd>Redirect the stdout of a remote process to a given, local, file. Default
is to use <code>lsh</code>'s stdout. Like <code>--stdin</code>, it is reset
after it is used.
<br><dt><code>--stderr</code>
<dd>Redirect the stdout of a remote process to a given, local, file.
Analogous to the <code>--stdout</code> option.
<br><dt><code>-g</code>
<dd>Remote peers, aka Global forwarding. This option applies to the
forwarding actions, i.e. <code>-L</code> and <code>-R</code>. By default, only
connections to the loopback interface, ip 127.0.0.1, are forwared. This
implies that only processes on the same machine can use the forwarded
tunnel directly. If the -g modifier is in effect, the forwarding party
will listen on <em>all</em> network interfaces.
</dl>
<p><hr>
Node:<a name="Verbosity%20options">Verbosity options</a>,
Previous:<a rel=previous href="#Action%20options">Action options</a>,
Up:<a rel=up href="#Invoking%20lsh">Invoking lsh</a>
<br>
<h2>Verbosity options</h2>
<p>These options determines what messages <code>lsh</code> writes on
its stderr.
<dl>
<br><dt><code>-q</code>
<dd>Quiet mode. Disables all messages and all questions, except password
prompts and fatal internal errors.
<br><dt><code>-v</code>
<dd>Verbose mode. Makes <code>lsh</code> a little more verbose. The intention is
to provide information that is useful for ordinary trouble shooting,
and makes sense also to those not familiar with <code>lsh</code> internals.
<br><dt><code>--trace</code>
<dd>Trace mode. Prints some internal information to aid tracking
<code>lsh</code>'s flow of control.
<br><dt><code>--debug.</code>
<dd>Debug mode. Dumps <em>a lot</em> of information, including dumps of all
sent and received packets. It tries to avoid dumping highly sensitive data,
such as private keys and the contents of <code>SSH_MSG_USERAUTH_REQUEST</code>
messages, but you should still use it with care.
</dl>
<p>Note that all these options are orthogonal. If you use <code>--trace</code>,
you usually want to add <code>-v</code> as well; <code>--trace</code> does not
do that automatically.
<p><hr>
Node:<a name="Invoking%20lshd">Invoking lshd</a>,
Next:<a rel=next href="#Terminology">Terminology</a>,
Previous:<a rel=previous href="#Invoking%20lsh">Invoking lsh</a>,
Up:<a rel=up href="#Top">Top</a>
<br>
<h1>Invoking <code>lshd</code></h1>
<p><a name="lshd-usage"></a>
<p><code>lshd</code> is a server that accepts connections from clients
speaking the Secure Shell Protocol. It is usually started automatically
when the systems boots, and runs with root privileges. However, it is
also possible to start <code>lshd</code> manually, and with user
privileges.
<p>There are currently no configuration files. Instead, command line options
are used to tell <code>lshd</code> what to do. Many options have <code>--foo</code>
and <code>--no-foo</code> variants. Options specifying the default behaviour
are not listed here.
<p>Some of the options are the shared with <code>lsh</code>. In particular, see
<a href="#Algorithm%20options">Algorithm options</a> and <a href="#Verbosity%20options">Verbosity options</a>.
<p>Options specific to the <code>lshd</code> server are:
<dl>
<br><dt><code>-p</code>
<dd>Port to listen to. The mandatory argument is a decimal port number or a
service name. Default is "ssh", usually port 22.
<br><dt><code>--interface</code>
<dd>Network interface to listen on. By default, <code>lshd</code> listens on all
interfaces.
<br><dt><code>-h</code>
<dd>Location of the server's private key file. By default,
<code>/etc/lsh_host_key</code>.
<br><dt><code>-i</code>
<dd>Variant of the s-expression syntax to use when reading the host key.
Default is to use transport format. Not a terribly useful option.
<br><dt><code>--ssh1-fallback</code>
<dd>This options enables fallback to <code>ssh1</code>. <code>lshd</code> doesn't
implement version 1 of the Secure Shell Protocol. But it can fork an
<code>ssh1</code> server when an old client connects. Falling back to
<code>ssh1</code> is inefficient, and requires some special features of the
server fallen back to. It should work with the <code>sshd</code> daemon
supplied with reasonably new versions of F-Secure's <code>sshd1</code>,
and with OpenSSH.
<p>The optional argument provides the filename of the ssh1 daemon to use.
The default name is <code>/usr/local/sbin/sshd1</code>, unless something else
was configured at compile time.
<br><dt><code>--daemonic</code>
<dd>Enables daemonic mode. <code>lshd</code> forks into the background,
redirects its stdio file descriptors to <code>/dev/null</code>, changes its
working directory to <code>/</code>, and redirects any diagnostic or debugging
messages via syslog.
<p><code>lshd</code> should be able to deal with the environment it inherits
if it is started by <code>init</code> or <code>inetd</code>, but this is not
really tested.
<br><dt><code>--pid-file</code>
<dd>Creates a locked pid file, to make it easier to write start and stop
scripts for <code>lshd</code>. The mandatory argument provides the filename.
This option is enabled by default when operating in daemonic mode, and
the default filename is <code>/var/run/lshd.pid</code>.
<br><dt><code>--enable-core</code>
<dd>By default, <code>lshd</code> disables core dumps, to avoid leaking sensitive
information. This option changes that behaviour, and allows <code>lshd</code>
to dump core on fatal errors.
<br><dt><code>--no-password</code>
<dd>Disable the "password" user authentication mechanism.
<br><dt><code>--no-publickey</code>
<dd>Disable the "publickey" user authentication mechanism.
<br><dt><code>--root-login</code>
<dd>Enable root login. By default, root can not log in using <code>lshd</code>.
<br><dt><code>--kerberos-passwords</code>
<dd>Verify passwords against the kerberos database. This is implemented
using the <code>lsh-krb-checkpw</code> helper program. Note that this does
<em>not</em> use the Kerberos infrastructure in the Right Way. Experimental.
<br><dt><code>--password-helper</code>
<dd>Tells <code>lshd</code> to use a helper program for verifying passwords.
This is a generalization of <code>--kerberos-passwords</code>, and it could
be used for verifying passwords against any password database. See the
source files <code>lsh-krb-checkpw.c</code> and <code>unix_user.c</code> for
details.
<br><dt><code>--login-shell</code>
<dd>Use the specified program as the login shell for all users, overriding
the login shell in the passwd database.
<br><dt><code>--srp-keyexchange</code>
<dd>Enable <small>SRP</small> keyexchange and user authentication.
<br><dt><code>--no-pty-support</code>
<dd>Disable support for pseudo terminals.
<br><dt><code>--no-tcp-forward</code>
<dd>Disable support for tcp forwarding, in both directions.
</dl>
<p><hr>
Node:<a name="Terminology">Terminology</a>,
Next:<a rel=next href="#Concept%20Index">Concept Index</a>,
Previous:<a rel=previous href="#Invoking%20lshd">Invoking lshd</a>,
Up:<a rel=up href="#Top">Top</a>
<br>
<h1>Terminology</h1>
<p><hr>
Node:<a name="Concept%20Index">Concept Index</a>,
Previous:<a rel=previous href="#Terminology">Terminology</a>,
Up:<a rel=up href="#Top">Top</a>
<br>
<h1>Concept Index</h1>
<ul compact>
</ul>
<h1>Table of Contents</h1>
<ul>
<li><a href="#Top"></a>
<li><a href="#Introduction">Introduction</a>
<ul>
<li><a href="#Threats">Threats</a>
<li><a href="#Features">Features</a>
<li><a href="#Related%20techniques">Related programs and techniques</a>
<ul>
<li><a href="#ssh1"><code>ssh-1.x</code></a>
<li><a href="#ssh2"><code>ssh-2.x</code></a>
<li><a href="#Kerberos">Kerberos</a>
<li><a href="#ipsec"><small>IPSEC</small></a>
</ul>
</ul>
<li><a href="#Installation">Installation</a>
<li><a href="#Getting%20started">Getting started</a>
<ul>
<li><a href="#lsh%20basics"><code>lsh</code> basics</a>
<li><a href="#tcpip%20forwarding">Port forwarding</a>
<li><a href="#lshd%20basics"><code>lshd</code> basics</a>
<li><a href="#public-key">Using public-key user authentication</a>
<li><a href="#srp">Using <small>SRP</small> authentication</a>
<li><a href="#sexp">Examining keys and other sexp files</a>
<li><a href="#Converting%20keys">Converting keys from <code>ssh2</code> and OpenSSH</a>
</ul>
<li><a href="#Invoking%20lsh">Invoking <code>lsh</code></a>
<ul>
<li><a href="#Algorithm%20options">Algorithm options</a>
<li><a href="#Hostauth%20options">Host authentication options</a>
<li><a href="#Userauth%20options">User authentication options</a>
<li><a href="#Action%20options">Action options</a>
<li><a href="#Verbosity%20options">Verbosity options</a>
</ul>
<li><a href="#Invoking%20lshd">Invoking <code>lshd</code></a>
<li><a href="#Terminology">Terminology</a>
<li><a href="#Concept%20Index">Concept Index</a>
</ul>
</body></html>