\C{using} Using PuTTY

This chapter provides a general introduction to some more advanced
features of PuTTY. For extreme detail and reference purposes,
\k{config} is likely to contain more information.

\H{using-session} During your session

A lot of PuTTY's complexity and features are in the configuration
panel. Once you have worked your way through that and started
a session, things should be reasonably simple after that.
Nevertheless, there are a few more useful features available.

\S{using-selection} Copying and pasting text

\I{copy and paste}Often in a PuTTY session you will find text on
your terminal screen which you want to type in again. Like most
other terminal emulators, PuTTY allows you to copy and paste the
text rather than having to type it again. Also, copy and paste uses
the \I{Windows clipboard}Windows \i{clipboard}, so that you can
paste (for example) URLs into a web browser, or paste from a word
processor or spreadsheet into your terminal session.

By default, PuTTY's copy and paste works entirely with the \i{mouse}.
(This will be familiar to people who have used \i\c{xterm} on Unix.)
In order to copy text to the clipboard, you just click the \i{left
mouse button} in the \i{terminal window}, and drag to
\I{selecting text}select text. When you let go of the button, the text
is \e{automatically} copied to the clipboard. You do not need to press
\i{Ctrl-C} or \i{Ctrl-Ins}; in fact, if you do press Ctrl-C, PuTTY will
send a Ctrl-C character down your session to the server where it will
probably cause a process to be interrupted.

Pasting into PuTTY is done using the right button (or the middle mouse
button, if you have a \i{three-button mouse} and have set it up; see
\k{config-mouse}). (Pressing \i{Shift-Ins}, or selecting \q{Paste}
from the \I{right mouse button, with Ctrl}Ctrl+right-click
\i{context menu}, have the same effect.) When
you click the \i{right mouse button}, PuTTY will read whatever is in
the Windows clipboard and paste it into your session. By default, this
behaves \e{exactly} as if the clipboard contents had been typed at the
keyboard; therefore, be careful of pasting formatted text into an
editor that does automatic \i{indenting}, as you may find that the spaces
pasted from the clipboard plus the spaces added by the editor add up
to too many spaces and ruin the formatting. (Some remote applications
can ask PuTTY to identify text that is being pasted, to avoid this
sort of problem; but if your application does not, there is nothing
PuTTY can do to avoid this.)

If you \i{double-click} the left mouse button, PuTTY will
\I{selecting words}select a whole word. If you double-click, hold
down the second click, and drag the mouse, PuTTY will select a
sequence of whole words. (You can adjust precisely what PuTTY
considers to be part of a word; see \k{config-charclasses}.)
If you \e{triple}-click, or \i{triple-click} and drag, then
PuTTY will \I{selecting lines}select a whole line or sequence of lines.

If you want to select a \I{rectangular selection}rectangular region
instead of selecting to the end of each line, you can do this by
holding down Alt when you make your selection. You can also
configure rectangular selection to be the default, and then holding
down Alt gives the normal behaviour instead: see
\k{config-rectselect} for details.

(In some Unix environments, Alt+drag is intercepted by the window
manager. Shift+Alt+drag should work for rectangular selection as
well, so you could try that instead.)

If you have a \i{middle mouse button}, then you can use it to
\I{adjusting a selection}adjust an existing selection if you
selected something slightly wrong. (If you have configured the
middle mouse button to paste, then the right mouse button does this
instead.) Click the button on the screen, and you can pick up the
nearest end of the selection and drag it to somewhere else.

If you are running PuTTY itself on Unix (not just using it to connect
to a Unix system from Windows), by default you will likely have to use
similar mouse actions in other applications to paste the text you
copied from PuTTY, and to copy text for pasting into PuTTY; actions
like \i{Ctrl-C} and Ctrl-V will likely not behave as you expect.
\K{config-clipboards} explains why this is, and how you can change the
behaviour. (On Windows there is only a single selection shared with other
applications, so this confusion does not arise.)

It's possible for the server to ask to \I{mouse reporting}handle mouse
clicks in the PuTTY window itself.  If this happens, the \i{mouse pointer}
will turn into an arrow, and using the mouse to copy and paste will only
work if you hold down Shift.  See \k{config-features-mouse} and
\k{config-mouseshift} for details of this feature and how to configure
it.

You can customise much of this behaviour, for instance to enable copy
and paste from the keyboard; see \k{config-selection}.

\S{using-scrollback} \I{scrollback}Scrolling the screen back

PuTTY keeps track of text that has scrolled up off the top of the
terminal. So if something appears on the screen that you want to
read, but it scrolls too fast and it's gone by the time you try to
look for it, you can use the \i{scrollbar} on the right side of the
window to look back up the session \i{history} and find it again.

As well as using the scrollbar, you can also page the scrollback up
and down by pressing \i{Shift-PgUp} and \i{Shift-PgDn}. You can
scroll a line at a time using \i{Ctrl-PgUp} and \i{Ctrl-PgDn}, or
to the top/bottom of the scrollback with \i{Ctrl-Shift-PgUp} and
\i{Ctrl-Shift-PgDn}. These are still available if you configure the
scrollbar to be invisible.

By default the last 2000 lines scrolled off the top are
preserved for you to look at. You can increase (or decrease) this
value using the configuration box; see \k{config-scrollback}.

\S{using-sysmenu} The \ii{System menu}

If you click the left mouse button on the icon in the top left
corner of PuTTY's terminal window, or click the right mouse button
on the title bar, you will see the standard Windows system menu
containing items like Minimise, Move, Size and Close.

PuTTY's system menu contains extra program features in addition to
the Windows standard options. These extra menu commands are
described below.

(These options are also available in a \i{context menu} brought up
by holding Ctrl and clicking with the right mouse button anywhere
in the \i{PuTTY window}.)

\S2{using-eventlog} The PuTTY \i{Event Log}

If you choose \q{Event Log} from the system menu, a small window
will pop up in which PuTTY logs significant events during the
connection. Most of the events in the log will probably take place
during session startup, but a few can occur at any point in the
session, and one or two occur right at the end.

You can use the mouse to select one or more lines of the Event Log,
and hit the Copy button to copy them to the \i{clipboard}. If you
are reporting a bug, it's often useful to paste the contents of the
Event Log into your bug report.

(The Event Log is not the same as the facility to create a log file
of your session; that's described in \k{using-logging}.)

\S2{using-specials} \ii{Special commands}

Depending on the protocol used for the current session, there may be
a submenu of \q{special commands}. These are protocol-specific
tokens, such as a \q{break} signal, that can be sent down a
connection in addition to normal data. Their precise effect is usually
up to the server. Currently only Telnet, SSH, and serial connections
have special commands.

The \q{break} signal can also be invoked from the keyboard with
\i{Ctrl-Break}.

In an SSH connection, the following \I{SSH special commands}special
commands are available:

\b \I{IGNORE message, SSH special command}\I{No-op, in SSH}\ii{IGNORE message}

\lcont{
Should have no effect.
}

\b \I{Repeat key exchange, SSH special command}Repeat key exchange

\lcont{
Only available in SSH-2. Forces a \i{repeat key exchange} immediately (and
resets associated timers and counters). For more information about
repeat key exchanges, see \k{config-ssh-kex-rekey}.
}

\b \I{host key cache}Cache new host key type

\lcont{
Only available in SSH-2. This submenu appears only if the server has
host keys of a type that PuTTY doesn't already have cached, and so
won't consider. Selecting a key here will allow PuTTY to use that key
now and in future: PuTTY will do a fresh key-exchange with the selected
key, and immediately add that key to its permanent cache (relying on
the host key used at the start of the connection to cross-certify the
new key). That key will be used for the rest of the current session;
it may not actually be used for future sessions, depending on your
preferences (see \k{config-ssh-hostkey-order}).

Normally, PuTTY will carry on using a host key it already knows, even
if the server offers key formats that PuTTY would otherwise prefer,
to avoid host key prompts. As a result, if you've been using a server
for some years, you may still be using an older key than a new user
would use, due to server upgrades in the meantime. The SSH protocol
unfortunately does not have organised facilities for host key migration
and rollover, but this allows you to \I{host keys, upgrading}manually
upgrade.
}

\b \I{Break, SSH special command}Break

\lcont{
Only available in SSH-2, and only during a session. Optional
extension; may not be supported by server. PuTTY requests the server's
default break length.
}

\b \I{Signal, SSH special command}Signals (SIGINT, SIGTERM etc)

\lcont{
Only available in SSH-2, and only during a session. Sends various
POSIX signals. Not honoured by all servers.
}

The following \I{Telnet special commands}special commands are
available in Telnet:

\b \I{Are You There, Telnet special command}Are You There

\b \I{Break, Telnet special command}Break

\b \I{Synch, Telnet special command}Synch

\b \I{Erase Character, Telnet special command}Erase Character

\lcont{
PuTTY can also be configured to send this when the Backspace key is
pressed; see \k{config-telnetkey}.
}

\b \I{Erase Line, Telnet special command}Erase Line

\b \I{Go Ahead, Telnet special command}Go Ahead

\b \I{No Operation, Telnet special command}No Operation

\lcont{
Should have no effect.
}

\b \I{Abort Process, Telnet special command}Abort Process

\b \I{Abort Output, Telnet special command}Abort Output

\b \I{Interrupt Process, Telnet special command}Interrupt Process

\lcont{
PuTTY can also be configured to send this when Ctrl-C is typed; see
\k{config-telnetkey}.
}

\b \I{Suspend Process, Telnet special command}Suspend Process

\lcont{
PuTTY can also be configured to send this when Ctrl-Z is typed; see
\k{config-telnetkey}.
}

\b \I{End Of Record, Telnet special command}End Of Record

\b \I{End Of File, Telnet special command}End Of File

With a serial connection, the only available special command is
\I{Break, serial special command}\q{Break}.

\S2{using-newsession} Starting new sessions

PuTTY's system menu provides some shortcut ways to start new
sessions:

\b Selecting \i{\q{New Session}} will start a completely new
instance of PuTTY, and bring up the configuration box as normal.

\b Selecting \i{\q{Duplicate Session}} will start a session in a
new window with precisely the same options as your current one -
connecting to the same host using the same protocol, with all the
same terminal settings and everything.

\b In an inactive window, selecting \i{\q{Restart Session}} will
do the same as \q{Duplicate Session}, but in the current window.

\b The \i{\q{Saved Sessions} submenu} gives you quick access to any
sets of stored session details you have previously saved. See
\k{config-saving} for details of how to create saved sessions.

\S2{using-changesettings} \I{settings, changing}Changing your
session settings

If you select \i{\q{Change Settings}} from the system menu, PuTTY will
display a cut-down version of its initial configuration box. This
allows you to adjust most properties of your current session. You
can change the terminal size, the font, the actions of various
keypresses, the colours, and so on.

Some of the options that are available in the main configuration box
are not shown in the cut-down Change Settings box. These are usually
options which don't make sense to change in the middle of a session
(for example, you can't switch from SSH to Telnet in mid-session).

You can save the current settings to a saved session for future use
from this dialog box. See \k{config-saving} for more on saved
sessions.

\S2{using-copyall} \i{Copy All to Clipboard}

This system menu option provides a convenient way to copy the whole
contents of the terminal screen (up to the last nonempty line) and
scrollback to the \i{clipboard} in one go.

\S2{reset-terminal} \I{scrollback, clearing}Clearing and
\I{terminal, resetting}resetting the terminal

The \i{\q{Clear Scrollback}} option on the system menu tells PuTTY
to discard all the lines of text that have been kept after they
scrolled off the top of the screen. This might be useful, for
example, if you displayed sensitive information and wanted to make
sure nobody could look over your shoulder and see it. (Note that
this only prevents a casual user from using the scrollbar to view
the information; the text is not guaranteed not to still be in
PuTTY's memory.)

The \i{\q{Reset Terminal}} option causes a full reset of the
\i{terminal emulation}. A VT-series terminal is a complex piece of
software and can easily get into a state where all the text printed
becomes unreadable. (This can happen, for example, if you
accidentally output a binary file to your terminal.) If this
happens, selecting Reset Terminal should sort it out.

\S2{using-fullscreen} \ii{Full screen} mode

If you find the title bar on a maximised window to be ugly or
distracting, you can select Full Screen mode to maximise PuTTY
\q{even more}. When you select this, PuTTY will expand to fill the
whole screen and its borders, title bar and scrollbar will
disappear. (You can configure the scrollbar not to disappear in
full-screen mode if you want to keep it; see \k{config-scrollback}.)

When you are in full-screen mode, you can still access the \i{system
menu} if you click the left mouse button in the \e{extreme} top left
corner of the screen.

\H{using-logging} Creating a \i{log file} of your \I{session
log}session

For some purposes you may find you want to log everything that
appears on your screen. You can do this using the \q{Logging}
panel in the configuration box.

To begin a session log, select \q{Change Settings} from the system
menu and go to the Logging panel. Enter a log file name, and select
a logging mode. (You can log all session output including the
terminal \i{control sequence}s, or you can just log the printable text.
It depends what you want the log for.) Click \q{Apply} and your log
will be started. Later on, you can go back to the Logging panel and
select \q{Logging turned off completely} to stop logging; then PuTTY
will close the log file and you can safely read it.

See \k{config-logging} for more details and options.

\H{using-translation} Altering your \i{character set} configuration

If you find that special characters (\i{accented characters}, for
example, or \i{line-drawing characters}) are not being displayed
correctly in your PuTTY session, it may be that PuTTY is interpreting
the characters sent by the server according to the wrong \e{character
set}. There are a lot of different character sets available, and no
good way for PuTTY to know which to use, so it's entirely possible
for this to happen.

If you click \q{Change Settings} and look at the \q{Translation}
panel, you should see a large number of character sets which you can
select, and other related options. Now all you need is to find out
which of them you want! (See \k{config-translation} for more
information.)

\H{using-x-forwarding} Using \i{X11 forwarding} in SSH

The SSH protocol has the ability to securely forward X Window System
\i{graphical applications} over your encrypted SSH connection, so that
you can run an application on the SSH server machine and have it put
its windows up on your local machine without sending any X network
traffic in the clear.

In order to use this feature, you will need an X display server for
your Windows machine, such as Cygwin/X, X-Win32, or Exceed. This will probably
install itself as display number 0 on your local machine; if it
doesn't, the manual for the \i{X server} should tell you what it
does do.

You should then tick the \q{Enable X11 forwarding} box in the
X11 panel (see \k{config-ssh-x11}) before starting your SSH
session. The \i{\q{X display location}} box is blank by default, which
means that PuTTY will try to use a sensible default such as \c{:0},
which is the usual display location where your X server will be
installed. If that needs changing, then change it.

Now you should be able to log in to the SSH server as normal. To
check that X forwarding has been successfully negotiated during
connection startup, you can check the PuTTY Event Log (see
\k{using-eventlog}). It should say something like this:

\c 2001-12-05 17:22:01 Requesting X11 forwarding
\c 2001-12-05 17:22:02 X11 forwarding enabled

If the remote system is Unix or Unix-like, you should also be able
to see that the \i{\c{DISPLAY} environment variable} has been set to
point at display 10 or above on the SSH server machine itself:

\c fred@unixbox:~$ echo $DISPLAY
\c unixbox:10.0

If this works, you should then be able to run X applications in the
remote session and have them display their windows on your PC.

For more options relating to X11 forwarding, see \k{config-ssh-x11}.

\H{using-port-forwarding} Using \i{port forwarding} in SSH

The SSH protocol has the ability to forward arbitrary \I{network
connection}network (TCP) connections over your encrypted SSH
connection, to avoid the network traffic being sent in clear. For
example, you could use this to connect from your home computer to a
\i{POP-3} server on a remote machine without your POP-3 password being
visible to network sniffers.

In order to use port forwarding to \I{local port forwarding}connect
from your local machine to a port on a remote server, you need to:

\b Choose a \i{port number} on your local machine where PuTTY should
listen for incoming connections. There are likely to be plenty of
unused port numbers above 3000. (You can also use a local loopback
address here; see below for more details.)

\b Now, before you start your SSH connection, go to the Tunnels
panel (see \k{config-ssh-portfwd}). Make sure the \q{Local} radio
button is set. Enter the local port number into the \q{Source port}
box. Enter the destination host name and port number into the
\q{Destination} box, separated by a colon (for example,
\c{popserver.example.com:110} to connect to a POP-3 server).

\b Now click the \q{Add} button. The details of your port forwarding
should appear in the list box.

Now start your session and log in. (Port forwarding will not be
enabled until after you have logged in; otherwise it would be easy
to perform completely anonymous network attacks, and gain access to
anyone's virtual private network.) To check that PuTTY has set up
the port forwarding correctly, you can look at the PuTTY Event Log
(see \k{using-eventlog}). It should say something like this:

\c 2001-12-05 17:22:10 Local port 3110 forwarding to
\c          popserver.example.com:110

Now if you connect to the source port number on your local PC, you
should find that it answers you exactly as if it were the service
running on the destination machine. So in this example, you could
then configure an e-mail client to use \c{localhost:3110} as a POP-3
server instead of \c{popserver.example.com:110}. (Of course, the
forwarding will stop happening when your PuTTY session closes down.)

You can also forward ports in the other direction: arrange for a
particular port number on the \e{server} machine to be \I{remote
port forwarding}forwarded back to your PC as a connection to a
service on your PC or near it.
To do this, just select the \q{Remote} radio button instead of the
\q{Local} one. The \q{Source port} box will now specify a port
number on the \e{server} (note that most servers will not allow you
to use \I{privileged port}port numbers under 1024 for this purpose).

An alternative way to forward local connections to remote hosts is
to use \I{dynamic port forwarding}dynamic SOCKS proxying. In this
mode, PuTTY acts as a SOCKS server, which SOCKS-aware programs can
connect to and open forwarded connections to the destination of their
choice, so this can be an alternative to long lists of static
forwardings. To use this mode, you will need to select the \q{Dynamic}
radio button instead of \q{Local}, and then you should not enter
anything into the \q{Destination} box (it will be ignored). PuTTY will
then listen for SOCKS connections on the port you have specified.
Most \i{web browsers} can be configured to connect to this SOCKS proxy
service; also, you can forward other PuTTY connections through it by
setting up the Proxy control panel (see \k{config-proxy} for details).

The source port for a forwarded connection usually does not accept
connections from any machine except the \I{localhost}SSH client or
server machine itself (for local and remote forwardings respectively).
There are controls in the Tunnels panel to change this:

\b The \q{Local ports accept connections from other hosts} option
allows you to set up local-to-remote port forwardings (including
dynamic port forwardings) in such a way that machines other than
your client PC can connect to the forwarded port.

\b The \q{Remote ports do the same} option does the same thing for
remote-to-local port forwardings (so that machines other than the
SSH server machine can connect to the forwarded port.) Note that
this feature is only available in the SSH-2 protocol, and not all
SSH-2 servers honour it (in \i{OpenSSH}, for example, it's usually
disabled by default).

You can also specify an \i{IP address} to \I{listen address}listen
on. Typically a Windows machine can be asked to listen on any single
IP address in the \cw{127.*.*.*} range, and all of these are
\i{loopback address}es available only to the local machine. So if
you forward (for example) \c{127.0.0.5:79} to a remote machine's
\i\cw{finger} port, then you should be able to run commands such as
\c{finger fred@127.0.0.5}.
This can be useful if the program connecting to the forwarded port
doesn't allow you to change the port number it uses. This feature is
available for local-to-remote forwarded ports; SSH-1 is unable to
support it for remote-to-local ports, while SSH-2 can support it in
theory but servers will not necessarily cooperate.

(Note that if you're using Windows XP Service Pack 2, you may need
to obtain a fix from Microsoft in order to use addresses like
\cw{127.0.0.5} - see \k{faq-alternate-localhost}.)

For more options relating to port forwarding, see
\k{config-ssh-portfwd}.

If the connection you are forwarding over SSH is itself a second SSH
connection made by another copy of PuTTY, you might find the
\q{logical host name} configuration option useful to warn PuTTY of
which host key it should be expecting. See \k{config-loghost} for
details of this.

\H{using-rawprot} Making \i{raw TCP connections}

A lot of \I{debugging Internet protocols}Internet protocols are
composed of commands and responses in plain text. For example,
\i{SMTP} (the protocol used to transfer e-mail), \i{NNTP} (the
protocol used to transfer Usenet news), and \i{HTTP} (the protocol
used to serve Web pages) all consist of commands in readable plain
text.

Sometimes it can be useful to connect directly to one of these
services and speak the protocol \q{by hand}, by typing protocol
commands and watching the responses. On Unix machines, you can do
this using the system's \c{telnet} command to connect to the right
port number. For example, \c{telnet mailserver.example.com 25} might
enable you to talk directly to the SMTP service running on a mail
server.

Although the Unix \c{telnet} program provides this functionality,
the protocol being used is not really Telnet. Really there is no
actual protocol at all; the bytes sent down the connection are
exactly the ones you type, and the bytes shown on the screen are
exactly the ones sent by the server. Unix \c{telnet} will attempt to
detect or guess whether the service it is talking to is a real
Telnet service or not; PuTTY prefers to be told for certain.

In order to make a debugging connection to a service of this type,
you simply select the fourth protocol name, \I{\q{Raw}
protocol}\q{Raw}, from the \q{Protocol} buttons in the \q{Session}
configuration panel. (See \k{config-hostname}.) You can then enter a
host name and a port number, and make the connection.

\H{using-serial} Connecting to a local serial line

PuTTY can connect directly to a local serial line as an alternative
to making a network connection. In this mode, text typed into the
PuTTY window will be sent straight out of your computer's serial
port, and data received through that port will be displayed in the
PuTTY window. You might use this mode, for example, if your serial
port is connected to another computer which has a serial connection.

To make a connection of this type, simply select \q{Serial} from the
\q{Connection type} radio buttons on the \q{Session} configuration
panel (see \k{config-hostname}). The \q{Host Name} and \q{Port}
boxes will transform into \q{Serial line} and \q{Speed}, allowing
you to specify which serial line to use (if your computer has more
than one) and what speed (baud rate) to use when transferring data.
For further configuration options (data bits, stop bits, parity,
flow control), you can use the \q{Serial} configuration panel (see
\k{config-serial}).

After you start up PuTTY in serial mode, you might find that you
have to make the first move, by sending some data out of the serial
line in order to notify the device at the other end that someone is
there for it to talk to. This probably depends on the device. If you
start up a PuTTY serial session and nothing appears in the window,
try pressing Return a few times and see if that helps.

A serial line provides no well defined means for one end of the
connection to notify the other that the connection is finished.
Therefore, PuTTY in serial mode will remain connected until you
close the window using the close button.

\H{using-supdup} Connecting using the SUPDUP protocol

PuTTY can use the SUPDUP protocol to connect to a server. SUPDUP is a
login protocol used mainly by PDP-10 and Lisp machines during the
period 1975-1990.  Like Telnet and Rlogin, it is unsecured, so modern
systems almost never support it.

To make a connection of this type, select \q{SUPDUP} from the
\q{Connection type} radio buttons on the \q{Session} panel (see
\k{config-hostname}). For further configuration options (character
set, more processing, scrolling), you can use the \q{SUPDUP}
configuration panel (see \k{config-supdup}).

The SUPDUP protocol does not support changing the terminal dimensions,
so this capability is disabled during a SUPDUP session.

SUPDUP provides no well defined means for one end of the connection to
notify the other that the connection is finished.  Therefore, PuTTY in
SUPDUP mode will remain connected until you close the window using the
close button.

\H{using-cmdline} The PuTTY command line

PuTTY can be made to do various things without user intervention by
supplying \i{command-line arguments} (e.g., from a \i{command prompt
window}, or a \i{Windows shortcut}).

\S{using-cmdline-session} Starting a session from the command line

\I\c{-ssh}\I\c{-telnet}\I\c{-rlogin}\I\c{-raw}\I\c{-serial}These
options allow you to bypass the configuration window and launch
straight into a session.

To start a connection to a server called \c{host}:

\c putty.exe [-ssh | -telnet | -rlogin | -raw] [user@]host

If this syntax is used, settings are taken from the \i{Default Settings}
(see \k{config-saving}); \c{user} overrides these settings if
supplied. Also, you can specify a protocol, which will override the
default protocol (see \k{using-cmdline-protocol}).

For telnet sessions, the following alternative syntax is supported
(this makes PuTTY suitable for use as a URL handler for \i{telnet
URLs} in \i{web browsers}):

\c putty.exe telnet://host[:port]/

To start a connection to a serial port, e.g. COM1:

\c putty.exe -serial com1

In order to start an existing saved session called \c{sessionname},
use the \c{-load} option (described in \k{using-cmdline-load}).

\c putty.exe -load "session name"

\S{using-cleanup} \i\c{-cleanup}

If invoked with the \c{-cleanup} option, rather than running as
normal, PuTTY will remove its \I{removing registry entries}registry
entries and \i{random seed file} from the local machine (after
confirming with the user). It will also attempt to remove information
about recently launched sessions stored in the \q{jump list} on
Windows 7 and up.

Note that on \i{multi-user systems}, \c{-cleanup} only removes
registry entries and files associated with the currently logged-in
user.

\S{using-general-opts} Standard command-line options

PuTTY and its associated tools support a range of command-line
options, most of which are consistent across all the tools. This
section lists the available options in all tools. Options which are
specific to a particular tool are covered in the chapter about that
tool.

\S2{using-cmdline-load} \i\c{-load}: load a saved session

\I{saved sessions, loading from command line}The \c{-load} option
causes PuTTY to load configuration details out of a saved session.
If these details include a host name, then this option is all you
need to make PuTTY start a session.

You need double quotes around the session name if it contains spaces.

If you want to create a \i{Windows shortcut} to start a PuTTY saved
session, this is the option you should use: your shortcut should
call something like

\c d:\path\to\putty.exe -load "my session"

(Note that PuTTY itself supports an alternative form of this option,
for backwards compatibility. If you execute \i\c{putty @sessionname}
it will have the same effect as \c{putty -load "sessionname"}. With
the \c{@} form, no double quotes are required, and the \c{@} sign
must be the very first thing on the command line. This form of the
option is deprecated.)

\S2{using-cmdline-protocol} Selecting a protocol: \c{-ssh},
\c{-telnet}, \c{-rlogin}, \c{-raw} \c{-serial}

To choose which protocol you want to connect with, you can use one
of these options:

\b \i\c{-ssh} selects the SSH protocol.

\b \i\c{-telnet} selects the Telnet protocol.

\b \i\c{-rlogin} selects the Rlogin protocol.

\b \i\c{-raw} selects the raw protocol.

\b \i\c{-serial} selects a serial connection.

These options are not available in the file transfer tools PSCP and
PSFTP (which only work with the SSH protocol).

These options are equivalent to the \i{protocol selection} buttons
in the Session panel of the PuTTY configuration box (see
\k{config-hostname}).

\S2{using-cmdline-v} \i\c{-v}: increase verbosity

\I{verbose mode}Most of the PuTTY tools can be made to tell you more
about what they are doing by supplying the \c{-v} option. If you are
having trouble when making a connection, or you're simply curious,
you can turn this switch on and hope to find out more about what is
happening.

\S2{using-cmdline-l} \i\c{-l}: specify a \i{login name}

You can specify the user name to log in as on the remote server
using the \c{-l} option. For example, \c{plink login.example.com -l
fred}.

These options are equivalent to the username selection box in the
Connection panel of the PuTTY configuration box (see
\k{config-username}).

\S2{using-cmdline-portfwd} \I{-L-upper}\c{-L}, \I{-R-upper}\c{-R}
and \I{-D-upper}\c{-D}: set up \i{port forwardings}

As well as setting up port forwardings in the PuTTY configuration
(see \k{config-ssh-portfwd}), you can also set up forwardings on the
command line. The command-line options work just like the ones in
Unix \c{ssh} programs.

To \I{local port forwarding}forward a local port (say 5110) to a
remote destination (say \cw{popserver.example.com} port 110), you
can write something like one of these:

\c putty -L 5110:popserver.example.com:110 -load mysession
\c plink mysession -L 5110:popserver.example.com:110

To forward a \I{remote port forwarding}remote port to a local
destination, just use the \c{-R} option instead of \c{-L}:

\c putty -R 5023:mytelnetserver.myhouse.org:23 -load mysession
\c plink mysession -R 5023:mytelnetserver.myhouse.org:23

To \I{listen address}specify an IP address for the listening end of the
tunnel, prepend it to the argument:

\c plink -L 127.0.0.5:23:localhost:23 myhost

To set up \I{dynamic port forwarding}SOCKS-based dynamic port
forwarding on a local port, use the \c{-D} option. For this one you
only have to pass the port number:

\c putty -D 4096 -load mysession

For general information on port forwarding, see
\k{using-port-forwarding}.

These options are not available in the file transfer tools PSCP and
PSFTP.

\S2{using-cmdline-m} \i\c{-m}: \I{reading commands from a file}read
a remote command or script from a file

The \i\c{-m} option performs a similar function to the \q{\ii{Remote
command}} box in the SSH panel of the PuTTY configuration box (see
\k{config-command}). However, the \c{-m} option expects to be given
a local file name, and it will read a command from that file.

With some servers (particularly Unix systems), you can even put
multiple lines in this file and execute more than one command in
sequence, or a whole shell script; but this is arguably an abuse, and
cannot be expected to work on all servers. In particular, it is known
\e{not} to work with certain \q{embedded} servers, such as \i{Cisco}
routers.

This option is not available in the file transfer tools PSCP and
PSFTP.

\S2{using-cmdline-p} \I{-P-upper}\c{-P}: specify a \i{port number}

The \c{-P} option is used to specify the port number to connect to. If
you have a Telnet server running on port 9696 of a machine instead of
port 23, for example:

\c putty -telnet -P 9696 host.name
\c plink -telnet -P 9696 host.name

(Note that this option is more useful in Plink than in PuTTY,
because in PuTTY you can write \c{putty -telnet host.name 9696} in
any case.)

This option is equivalent to the port number control in the Session
panel of the PuTTY configuration box (see \k{config-hostname}).

\S2{using-cmdline-pw} \i\c{-pw}: specify a \i{password}

A simple way to automate a remote login is to supply your password
on the command line. This is \e{not recommended} for reasons of
security. If you possibly can, we recommend you set up public-key
authentication instead. See \k{pubkey} for details.

Note that the \c{-pw} option only works when you are using the SSH
protocol. Due to fundamental limitations of Telnet, Rlogin, and
SUPDUP, these protocols do not support automated password
authentication.

\S2{using-cmdline-agentauth} \i\c{-agent} and \i\c{-noagent}:
control use of Pageant for authentication

The \c{-agent} option turns on SSH authentication using Pageant, and
\c{-noagent} turns it off. These options are only meaningful if you
are using SSH.

See \k{pageant} for general information on \i{Pageant}.

These options are equivalent to the agent authentication checkbox in
the Auth panel of the PuTTY configuration box (see
\k{config-ssh-tryagent}).

\S2{using-cmdline-agent} \I{-A-upper}\c{-A} and \i\c{-a}: control \i{agent
forwarding}

The \c{-A} option turns on SSH agent forwarding, and \c{-a} turns it
off. These options are only meaningful if you are using SSH.

See \k{pageant} for general information on \i{Pageant}, and
\k{pageant-forward} for information on agent forwarding. Note that
there is a security risk involved with enabling this option; see
\k{pageant-security} for details.

These options are equivalent to the agent forwarding checkbox in the
Auth panel of the PuTTY configuration box (see \k{config-ssh-agentfwd}).

These options are not available in the file transfer tools PSCP and
PSFTP.

\S2{using-cmdline-x11} \I{-X-upper}\c{-X} and \i\c{-x}: control \i{X11
forwarding}

The \c{-X} option turns on X11 forwarding in SSH, and \c{-x} turns
it off. These options are only meaningful if you are using SSH.

For information on X11 forwarding, see \k{using-x-forwarding}.

These options are equivalent to the X11 forwarding checkbox in the
X11 panel of the PuTTY configuration box (see \k{config-ssh-x11}).

These options are not available in the file transfer tools PSCP and
PSFTP.

\S2{using-cmdline-pty} \i\c{-t} and \I{-T-upper}\c{-T}: control
\i{pseudo-terminal allocation}

The \c{-t} option ensures PuTTY attempts to allocate a
pseudo-terminal at the server, and \c{-T} stops it from allocating
one. These options are only meaningful if you are using SSH.

These options are equivalent to the \q{Don't allocate a
pseudo-terminal} checkbox in the SSH panel of the PuTTY
configuration box (see \k{config-ssh-pty}).

These options are not available in the file transfer tools PSCP and
PSFTP.

\S2{using-cmdline-noshell} \I{-N-upper}\c{-N}: suppress starting a
\I{suppressing remote shell}shell or command

The \c{-N} option prevents PuTTY from attempting to start a shell or
command on the remote server. You might want to use this option if
you are only using the SSH connection for port forwarding, and your
user account on the server does not have the ability to run a shell.

This feature is only available in SSH protocol version 2 (since the
version 1 protocol assumes you will always want to run a shell).

This option is equivalent to the \q{Don't start a shell or command
at all} checkbox in the SSH panel of the PuTTY configuration box
(see \k{config-ssh-noshell}).

This option is not available in the file transfer tools PSCP and
PSFTP.

\S2{using-cmdline-ncmode} \I{-nc}\c{-nc}: make a \i{remote network
connection} in place of a remote shell or command

The \c{-nc} option prevents Plink (or PuTTY) from attempting to
start a shell or command on the remote server. Instead, it will
instruct the remote server to open a network connection to a host
name and port number specified by you, and treat that network
connection as if it were the main session.

You specify a host and port as an argument to the \c{-nc} option,
with a colon separating the host name from the port number, like
this:

\c plink host1.example.com -nc host2.example.com:1234

You might want to use this feature if you needed to make an SSH
connection to a target host which you can only reach by going
through a proxy host, and rather than using port forwarding you
prefer to use the local proxy feature (see \k{config-proxy-type} for
more about local proxies). In this situation you might select
\q{Local} proxy type, set your local proxy command to be \cq{plink
%proxyhost -nc %host:%port}, enter the target host name on the
Session panel, and enter the directly reachable proxy host name on
the Proxy panel.

This feature is only available in SSH protocol version 2 (since the
version 1 protocol assumes you will always want to run a shell). It
is not available in the file transfer tools PSCP and PSFTP. It is
available in PuTTY itself, although it is unlikely to be very useful
in any tool other than Plink. Also, \c{-nc} uses the same server
functionality as port forwarding, so it will not work if your server
administrator has disabled port forwarding.

(The option is named \c{-nc} after the Unix program
\W{http://www.vulnwatch.org/netcat/}\c{nc}, short for \q{netcat}.
The command \cq{plink host1 -nc host2:port} is very similar in
functionality to \cq{plink host1 nc host2 port}, which invokes
\c{nc} on the server and tells it to connect to the specified
destination. However, Plink's built-in \c{-nc} option does not
depend on the \c{nc} program being installed on the server.)

\S2{using-cmdline-compress} \I{-C-upper}\c{-C}: enable \i{compression}

The \c{-C} option enables compression of the data sent across the
network. This option is only meaningful if you are using SSH.

This option is equivalent to the \q{Enable compression} checkbox in
the SSH panel of the PuTTY configuration box (see
\k{config-ssh-comp}).

\S2{using-cmdline-sshprot} \i\c{-1} and \i\c{-2}: specify an \i{SSH
protocol version}

The \c{-1} and \c{-2} options force PuTTY to use version \I{SSH-1}1
or version \I{SSH-2}2 of the SSH protocol. These options are only
meaningful if you are using SSH.

These options are equivalent to selecting the SSH protocol version in
the SSH panel of the PuTTY configuration box (see \k{config-ssh-prot}).

\S2{using-cmdline-ipversion} \i\c{-4} and \i\c{-6}: specify an
\i{Internet protocol version}

The \c{-4} and \c{-6} options force PuTTY to use the older Internet
protocol \i{IPv4} or the newer \i{IPv6} for most outgoing
connections.

These options are equivalent to selecting your preferred Internet
protocol version as \q{IPv4} or \q{IPv6} in the Connection panel of
the PuTTY configuration box (see \k{config-address-family}).

\S2{using-cmdline-identity} \i\c{-i}: specify an SSH \i{private key}

The \c{-i} option allows you to specify the name of a private key
file in \c{*.\i{PPK}} format which PuTTY will use to authenticate with the
server. This option is only meaningful if you are using SSH.

If you are using Pageant, you can also specify a \e{public} key file
(in RFC 4716 or OpenSSH format) to identify a specific key file to use.
(This won't work if you're not running Pageant, of course.)

For general information on \i{public-key authentication}, see
\k{pubkey}.

This option is equivalent to the \q{Private key file for
authentication} box in the Auth panel of the PuTTY configuration box
(see \k{config-ssh-privkey}).

\S2{using-cmdline-loghost} \i\c{-loghost}: specify a \i{logical host
name}

This option overrides PuTTY's normal SSH \I{host key cache}host key
caching policy by telling it the name of the host you expect your
connection to end up at (in cases where this differs from the location
PuTTY thinks it's connecting to). It can be a plain host name, or a
host name followed by a colon and a port number. See
\k{config-loghost} for more detail on this.

\S2{using-cmdline-hostkey} \i\c{-hostkey}: \I{manually configuring
host keys}manually specify an expected host key

This option overrides PuTTY's normal SSH \I{host key cache}host key
caching policy by telling it exactly what host key to expect, which
can be useful if the normal automatic host key store in the Registry
is unavailable. The argument to this option should be either a host key
fingerprint, or an SSH-2 public key blob. See
\k{config-ssh-kex-manual-hostkeys} for more information.

You can specify this option more than once if you want to configure
more than one key to be accepted.

\S2{using-cmdline-pgpfp} \i\c{-pgpfp}: display \i{PGP key fingerprint}s

This option causes the PuTTY tools not to run as normal, but instead
to display the fingerprints of the PuTTY PGP Master Keys, in order to
aid with \i{verifying new versions}. See \k{pgpkeys} for more information.

\S2{using-cmdline-sercfg} \i\c{-sercfg}: specify serial port
\i{configuration}

This option specifies the configuration parameters for the serial
port (baud rate, stop bits etc). Its argument is interpreted as a
comma-separated list of configuration options, which can be as
follows:

\b Any single digit from 5 to 9 sets the number of data bits.

\b \cq{1}, \cq{1.5} or \cq{2} sets the number of stop bits.

\b Any other numeric string is interpreted as a baud rate.

\b A single lower-case letter specifies the parity: \cq{n} for none,
\cq{o} for odd, \cq{e} for even, \cq{m} for mark and \cq{s} for space.

\b A single upper-case letter specifies the flow control: \cq{N} for
none, \cq{X} for XON/XOFF, \cq{R} for RTS/CTS and \cq{D} for
DSR/DTR.

For example, \cq{-sercfg 19200,8,n,1,N} denotes a baud rate of
19200, 8 data bits, no parity, 1 stop bit and no flow control.

\S2{using-cmdline-sshlog} \i\c{-sessionlog}, \i\c{-sshlog},
\i\c{-sshrawlog}: specify session logging

These options cause the PuTTY network tools to write out a \i{log
file}. Each of them expects a file name as an argument, e.g.
\cq{-sshlog putty.log} causes an SSH packet log to be written to a
file called \cq{putty.log}. The three different options select
different logging modes, all available from the GUI too:

\b \c{-sessionlog} selects \q{All session output} logging mode.

\b \c{-sshlog} selects \q{SSH packets} logging mode.

\b \c{-sshrawlog} selects \q{SSH packets and raw data} logging mode.

For more information on logging configuration, see \k{config-logging}.

\S2{using-cmdline-proxycmd} \i\c{-proxycmd}: specify a local proxy
command

This option enables PuTTY's mode for running a \I{Local proxy}command
on the local machine and using it as a proxy for the network
connection. It expects a shell command string as an argument.

See \k{config-proxy-type} for more information on this, and on other
proxy settings. In particular, note that since the special sequences
described there are understood in the argument string, literal
backslashes must be doubled (if you want \c{\\} in your command, you
must put \c{\\\\} on the command line).

\S2{using-cmdline-restrict-acl} \i\c{-restrict-acl}: restrict the
\i{Windows process ACL}

This option (on Windows only) causes PuTTY (or another PuTTY tool) to
try to lock down the operating system's access control on its own
process. If this succeeds, it should present an extra obstacle to
malware that has managed to run under the same user id as the PuTTY
process, by preventing it from attaching to PuTTY using the same
interfaces debuggers use and either reading sensitive information out
of its memory or hijacking its network session.

This option is not enabled by default, because this form of
interaction between Windows programs has many legitimate uses,
including accessibility software such as screen readers. Also, it
cannot provide full security against this class of attack in any case,
because PuTTY can only lock down its own ACL \e{after} it has started
up, and malware could still get in if it attacks the process between
startup and lockdown. So it trades away noticeable convenience, and
delivers less real security than you might want. However, if you do
want to make that tradeoff anyway, the option is available.

A PuTTY process started with \c{-restrict-acl} will pass that on to
any processes started with Duplicate Session, New Session etc.
(However, if you're invoking PuTTY tools explicitly, for instance as a
proxy command, you'll need to arrange to pass them the
\c{-restrict-acl} option yourself, if that's what you want.)

If Pageant is started with the \c{-restrict-acl} option, and you use
it to launch a PuTTY session from its System Tray submenu, then
Pageant will \e{not} default to starting the PuTTY subprocess with a
restricted ACL. This is because PuTTY is more likely to suffer reduced
functionality as a result of restricted ACLs (e.g. screen reader
software will have a greater need to interact with it), whereas
Pageant stores the more critical information (hence benefits more from
the extra protection), so it's reasonable to want to run Pageant but
not PuTTY with the ACL restrictions. You can force Pageant to start
subsidiary PuTTY processes with a restricted ACL if you also pass the
\i\c{-restrict-putty-acl} option.