1 \input texinfo @c -*-texinfo-*-
2 @c $Id: tinc.texi,v 1.8.4.45 2003/10/09 21:33:15 guus Exp $
9 @include tincinclude.texi
12 @dircategory Networking tools
14 * tinc: (tinc). The tinc Manual.
17 This is the info manual for @value{PACKAGE} version @value{VERSION}, a Virtual Private Network daemon.
19 Copyright @copyright{} 1998-2003 Ivo Timmermans
20 <ivo@@o2w.nl>, Guus Sliepen <guus@@sliepen.eu.org> and
21 Wessel Dankers <wsl@@nl.linux.org>.
23 $Id: tinc.texi,v 1.8.4.45 2003/10/09 21:33:15 guus Exp $
25 Permission is granted to make and distribute verbatim copies of this
26 manual provided the copyright notice and this permission notice are
27 preserved on all copies.
29 Permission is granted to copy and distribute modified versions of this
30 manual under the conditions for verbatim copying, provided that the
31 entire resulting derived work is distributed under the terms of a
32 permission notice identical to this one.
38 @subtitle Setting up a Virtual Private Network with tinc
39 @author Ivo Timmermans and Guus Sliepen
42 @vskip 0pt plus 1filll
44 This is the info manual for @value{PACKAGE} version @value{VERSION}, a Virtual Private Network daemon.
46 Copyright @copyright{} 1998-2003 Ivo Timmermans
47 <ivo@@o2w.nl>, Guus Sliepen <guus@@sliepen.eu.org> and
48 Wessel Dankers <wsl@@nl.linux.org>.
50 $Id: tinc.texi,v 1.8.4.45 2003/10/09 21:33:15 guus Exp $
52 Permission is granted to make and distribute verbatim copies of this
53 manual provided the copyright notice and this permission notice are
54 preserved on all copies.
56 Permission is granted to copy and distribute modified versions of this
57 manual under the conditions for verbatim copying, provided that the
58 entire resulting derived work is distributed under the terms of a
59 permission notice identical to this one.
64 @c ==================================================================
74 * Technical information::
75 * Platform specific information::
77 * Concept Index:: All used terms explained
81 @c ==================================================================
86 Tinc is a Virtual Private Network (VPN) daemon that uses tunneling and
87 encryption to create a secure private network between hosts on the
90 Because the tunnel appears to the IP level network code as a normal
91 network device, there is no need to adapt any existing software.
92 The encrypted tunnels allows VPN sites to share information with each other
93 over the Internet without exposing any information to others.
95 This document is the manual for tinc. Included are chapters on how to
96 configure your computer to use tinc, as well as the configuration
97 process of tinc itself.
100 * Virtual Private Networks::
102 * Supported platforms::
105 @c ==================================================================
106 @node Virtual Private Networks
107 @section Virtual Private Networks
110 A Virtual Private Network or VPN is a network that can only be accessed
111 by a few elected computers that participate. This goal is achievable in
112 more than just one way.
115 Private networks can consist of a single stand-alone Ethernet LAN. Or
116 even two computers hooked up using a null-modem cable. In these cases,
118 obvious that the network is @emph{private}, no one can access it from the
119 outside. But if your computers are linked to the Internet, the network
120 is not private anymore, unless one uses firewalls to block all private
121 traffic. But then, there is no way to send private data to trusted
122 computers on the other end of the Internet.
125 This problem can be solved by using @emph{virtual} networks. Virtual
126 networks can live on top of other networks, but they use encapsulation to
127 keep using their private address space so they do not interfere with
128 the Internet. Mostly, virtual networks appear like a singe LAN, even though
129 they can span the entire world. But virtual networks can't be secured
130 by using firewalls, because the traffic that flows through it has to go
131 through the Internet, where other people can look at it.
133 As is the case with either type of VPN, anybody could eavesdrop. Or
134 worse, alter data. Hence it's probably advisable to encrypt the data
135 that flows over the network.
137 When one introduces encryption, we can form a true VPN. Other people may
138 see encrypted traffic, but if they don't know how to decipher it (they
139 need to know the key for that), they cannot read the information that flows
140 through the VPN. This is what tinc was made for.
143 @c ==================================================================
148 I really don't quite remember what got us started, but it must have been
149 Guus' idea. He wrote a simple implementation (about 50 lines of C) that
150 used the ethertap device that Linux knows of since somewhere
151 about kernel 2.1.60. It didn't work immediately and he improved it a
152 bit. At this stage, the project was still simply called "vpnd".
154 Since then, a lot has changed---to say the least.
157 Tinc now supports encryption, it consists of a single daemon (tincd) for
158 both the receiving and sending end, it has become largely
159 runtime-configurable---in short, it has become a full-fledged
160 professional package.
162 @cindex Traditional VPNs
164 Tinc also allows more than two sites to connect to eachother and form a single VPN.
165 Traditionally VPNs are created by making tunnels, which only have two endpoints.
166 Larger VPNs with more sites are created by adding more tunnels.
167 Tinc takes another approach: only endpoints are specified,
168 the software itself will take care of creating the tunnels.
169 This allows for easier configuration and improved scalability.
171 A lot can---and will be---changed. We have a number of things that we would like to
172 see in the future releases of tinc. Not everything will be available in
173 the near future. Our first objective is to make tinc work perfectly as
174 it stands, and then add more advanced features.
176 Meanwhile, we're always open-minded towards new ideas. And we're
180 @c ==================================================================
181 @node Supported platforms
182 @section Supported platforms
185 Tinc has been verified to work under Linux, FreeBSD, OpenBSD, NetBSD, MacOS/X (Darwin), Solaris, and Windows (both natively and in a Cygwin environment),
186 with various hardware architectures. These are some of the platforms
187 that are supported by the universal tun/tap device driver or other virtual network device drivers.
188 Without such a driver, tinc will most
189 likely compile and run, but it will not be able to send or receive data
193 For an up to date list of supported platforms, please check the list on
195 @uref{http://tinc.nl.linux.org/platforms}.
198 @c ==================================================================
202 Tinc was first written for Linux running on an intel x86 processor, so
203 this is the best supported platform. The protocol however, and actually
204 anything about tinc, has been rewritten to support random byte ordering
205 and arbitrary word length. So in theory it should run on other
206 processors that Linux runs on. It has already been verified to run on
207 alpha and sparc processors as well.
209 Tinc uses the ethertap device or the universal tun/tap driver. The former is provided in the standard kernel
210 from version 2.1.60 up to 2.3.x, but has been replaced in favour of the tun/tap driver in kernel versions 2.4.0 and later.
213 @c ==================================================================
217 Tinc on FreeBSD relies on the universal tun/tap driver for its data
218 acquisition from the kernel. Therefore, tinc will work on the same platforms
219 as this driver. These are: FreeBSD 3.x, 4.x, 5.x.
222 @c ==================================================================
226 Tinc on OpenBSD relies on the tun driver for its data
227 acquisition from the kernel. It has been verified to work under at least OpenBSD 2.9.
229 Tunneling IPv6 packets may not work on OpenBSD.
232 @c ==================================================================
235 @c ==================================================================
239 Tinc on NetBSD relies on the tun driver for its data
240 acquisition from the kernel. It has been verified to work under at least NetBSD 1.5.2.
242 Tunneling IPv6 does not work on OpenBSD.
245 @c ==================================================================
249 Tinc on Solaris relies on the universal tun/tap driver for its data
250 acquisition from the kernel. Therefore, tinc will work on the same platforms
251 as this driver. It has been verified to work under Solaris 8 (SunOS 5.8).
253 IPv6 packets cannot be tunneled on Solaris.
255 @c ==================================================================
256 @subsection Darwin (MacOS/X)
260 Tinc on Darwin relies on the tunnel driver for its data
261 acquisition from the kernel. This driver is not part of Darwin but can be
262 downloaded from @uref{http://chrisp.de/en/projects/tunnel.html}.
264 IPv6 packets cannot be tunneled on Darwin.
266 @c ==================================================================
270 Tinc on Windows, in a Cygwin environment, relies on the CIPE driver or the TAP-Win32 driver for its data
271 acquisition from the kernel. This driver is not part of Windows but can be
272 downloaded from @uref{http://cipe-win32.sourceforge.net/}.
281 @c Preparing your system
288 @c ==================================================================
290 @chapter Preparations
292 This chapter contains information on how to prepare your system to
296 * Configuring the kernel::
301 @c ==================================================================
302 @node Configuring the kernel
303 @section Configuring the kernel
310 If you are running Linux, chances are good that your kernel already supports
311 all the devices that tinc needs for proper operation. For example, the
312 standard kernel from Redhat Linux already has support for ethertap and netlink
313 compiled in. Debian users can use the modconf utility to select the modules.
314 If your Linux distribution supports this method of selecting devices, look out
315 for something called `ethertap', and `netlink_dev' if it is using a kernel
316 version prior to 2.4.0. In that case you will need both these devices. If you
317 are using kernel 2.4.0 or later, you need to select `tun'.
320 If you can install these devices in a similar manner, you may skip this section.
321 Otherwise, you will have to recompile the kernel in order to turn on the required features.
322 If you are unfamiliar with the process of configuring and compiling a new kernel,
323 you should read the @uref{http://howto.linuxberg.com/LDP/HOWTO/Kernel-HOWTO.html, Kernel HOWTO} first.
326 * Configuration of Linux kernels 2.1.60 up to 2.4.0::
327 * Configuration of Linux kernels 2.4.0 and higher::
328 * Configuration of FreeBSD kernels::
329 * Configuration of OpenBSD kernels::
330 * Configuration of NetBSD kernels::
331 * Configuration of Solaris kernels::
332 * Configuration of Darwin (MacOS/X) kernels::
333 * Configuration of Windows::
337 @c ==================================================================
338 @node Configuration of Linux kernels 2.1.60 up to 2.4.0
339 @subsection Configuration of Linux kernels 2.1.60 up to 2.4.0
341 Here are the options you have to turn on when configuring a new kernel:
344 Code maturity level options
345 [*] Prompt for development and/or incomplete code/drivers
347 [*] Kernel/User netlink socket
348 <M> Netlink device emulation
349 Network device support
350 <M> Ethertap network tap
353 If you want to run more than one instance of tinc or other programs that use
354 the ethertap, you have to compile the ethertap driver as a module, otherwise
355 you can also choose to compile it directly into the kernel.
357 If you decide to build any of these as dynamic kernel modules, it's a good idea
358 to add these lines to @file{/etc/modules.conf}:
361 alias char-major-36 netlink_dev
363 options tap0 -o tap0 unit=0
365 options tap1 -o tap1 unit=1
367 alias tap@emph{N} ethertap
368 options tap@emph{N} -o tap@emph{N} unit=@emph{N}
371 Add as much alias/options lines as necessary.
374 @c ==================================================================
375 @node Configuration of Linux kernels 2.4.0 and higher
376 @subsection Configuration of Linux kernels 2.4.0 and higher
378 Here are the options you have to turn on when configuring a new kernel:
381 Code maturity level options
382 [*] Prompt for development and/or incomplete code/drivers
383 Network device support
384 <M> Universal tun/tap device driver support
387 It's not necessary to compile this driver as a module, even if you are going to
388 run more than one instance of tinc.
390 If you have an early 2.4 kernel, you can choose both the tun/tap driver and the
391 `Ethertap network tap' device. This latter is marked obsolete, and chances are
392 that it won't even function correctly anymore. Make sure you select the
393 universal tun/tap driver.
395 If you decide to build the tun/tap driver as a kernel module, add these lines
396 to @file{/etc/modules.conf}:
399 alias char-major-10-200 tun
403 @c ==================================================================
404 @node Configuration of FreeBSD kernels
405 @subsection Configuration of FreeBSD kernels
407 For FreeBSD version 4.1 and higher, the tap driver is included in the default kernel configuration, for earlier
408 systems (4.0 and earlier), you need to install the universal tun/tap driver
412 @c ==================================================================
413 @node Configuration of OpenBSD kernels
414 @subsection Configuration of OpenBSD kernels
416 For OpenBSD version 2.9 and higher,
417 the tun driver is included in the default kernel configuration.
420 @c ==================================================================
421 @node Configuration of NetBSD kernels
422 @subsection Configuration of NetBSD kernels
424 For NetBSD version 1.5.2 and higher,
425 the tun driver is included in the default kernel configuration.
428 @c ==================================================================
429 @node Configuration of Solaris kernels
430 @subsection Configuration of Solaris kernels
432 For Solaris 8 (SunOS 5.8) and higher,
433 the tun driver may or may not be included in the default kernel configuration.
434 If it isn't, the source can be downloaded from @uref{http://vtun.sourceforge.net/tun/}.
435 For x86 and sparc64 architectures, precompiled versions can be found at @uref{http://www.monkey.org/~dugsong/fragroute/}.
436 If the @file{net/if_tun.h} header file is missing, install it from the source package.
439 @c ==================================================================
440 @node Configuration of Darwin (MacOS/X) kernels
441 @subsection Configuration of Darwin (MacOS/X) kernels
443 Darwin does not come with a tunnel driver. You must download it at
444 @uref{http://chrisp.de/en/projects/tunnel.html}. If compiling the source fails,
445 try the binary module. The tunnel driver must be loaded before starting tinc
446 with the following command:
452 Once loaded, the tunnel driver will automatically create @file{/dev/tun0}..@file{/dev/tun3}
453 and the corresponding network interfaces.
456 @c ==================================================================
457 @node Configuration of Windows
458 @subsection Configuration of Windows
460 You will need to install the CIPE-Win32 driver or the TAP-Win32 driver, it
461 doesn't matter which one. You can download the CIPE driver from
462 @uref{http://cipe-win32.sourceforge.net}. Using the Network Connections
463 control panel, configure the CIPE-Win32 or TAP-Win32 network interface in the same way as you would
464 do from the tinc-up script as explained in the rest of the documentation.
467 @c ==================================================================
473 Before you can configure or build tinc, you need to have the OpenSSL,
474 zlib and lzo libraries installed on your system. If you try to configure tinc without
475 having them installed, configure will give you an error message, and stop.
484 @c ==================================================================
489 For all cryptography-related functions, tinc uses the functions provided
490 by the OpenSSL library.
492 If this library is not installed, you wil get an error when configuring
493 tinc for build. Support for running tinc without having OpenSSL
494 installed @emph{may} be added in the future.
496 You can use your operating system's package manager to install this if
497 available. Make sure you install the development AND runtime versions
500 If you have to install OpenSSL manually, you can get the source code
501 from @url{http://www.openssl.org/}. Instructions on how to configure,
502 build and install this package are included within the package. Please
503 make sure you build development and runtime libraries (which is the
506 If you installed the OpenSSL libraries from source, it may be necessary
507 to let configure know where they are, by passing configure one of the
508 --with-openssl-* parameters.
511 --with-openssl=DIR OpenSSL library and headers prefix
512 --with-openssl-include=DIR OpenSSL headers directory
513 (Default is OPENSSL_DIR/include)
514 --with-openssl-lib=DIR OpenSSL library directory
515 (Default is OPENSSL_DIR/lib)
519 @subsubheading License
522 The complete source code of tinc is covered by the GNU GPL version 2.
523 Since the license under which OpenSSL is distributed is not directly
524 compatible with the terms of the GNU GPL
525 @uref{http://www.openssl.org/support/faq.html#LEGAL2}, we
526 include an exemption to the GPL (see also the file COPYING.README) to allow
527 everyone to create a statically or dynamically linked executable:
530 This program is released under the GPL with the additional exemption
531 that compiling, linking, and/or using OpenSSL is allowed. You may
532 provide binary packages linked to the OpenSSL libraries, provided that
533 all other requirements of the GPL are met.
536 Since the LZO library used by tinc is also covered by the GPL,
537 we also present the following exemption:
540 Hereby I grant a special exception to the tinc VPN project
541 (http://tinc.nl.linux.org/) to link the LZO library with the OpenSSL library
542 (http://www.openssl.org).
544 Markus F.X.J. Oberhumer
548 @c ==================================================================
553 For the optional compression of UDP packets, tinc uses the functions provided
556 If this library is not installed, you wil get an error when configuring
557 tinc for build. Support for running tinc without having zlib
558 installed @emph{may} be added in the future.
560 You can use your operating system's package manager to install this if
561 available. Make sure you install the development AND runtime versions
564 If you have to install zlib manually, you can get the source code
565 from @url{http://www.gzip.org/zlib/}. Instructions on how to configure,
566 build and install this package are included within the package. Please
567 make sure you build development and runtime libraries (which is the
571 @c ==================================================================
576 Another form of compression is offered using the lzo library.
578 If this library is not installed, you wil get an error when configuring
579 tinc for build. Support for running tinc without having lzo
580 installed @emph{may} be added in the future.
582 You can use your operating system's package manager to install this if
583 available. Make sure you install the development AND runtime versions
586 If you have to install lzo manually, you can get the source code
587 from @url{http://www.oberhumer.com/opensource/lzo/}. Instructions on how to configure,
588 build and install this package are included within the package. Please
589 make sure you build development and runtime libraries (which is the
602 @c ==================================================================
604 @chapter Installation
606 If you use Debian, you may want to install one of the
607 precompiled packages for your system. These packages are equipped with
608 system startup scripts and sample configurations.
610 If you cannot use one of the precompiled packages, or you want to compile tinc
611 for yourself, you can use the source. The source is distributed under
612 the GNU General Public License (GPL). Download the source from the
613 @uref{http://tinc.nl.linux.org/download, download page}, which has
614 the checksums of these files listed; you may wish to check these with
615 md5sum before continuing.
617 Tinc comes in a convenient autoconf/automake package, which you can just
618 treat the same as any other package. Which is just untar it, type
619 `./configure' and then `make'.
620 More detailed instructions are in the file @file{INSTALL}, which is
621 included in the source distribution.
624 * Building and installing tinc::
629 @c ==================================================================
630 @node Building and installing tinc
631 @section Building and installing tinc
633 Detailed instructions on configuring the source, building tinc and installing tinc
634 can be found in the file called @file{INSTALL}.
636 @cindex binary package
637 If you happen to have a binary package for tinc for your distribution,
638 you can use the package management tools of that distribution to install tinc.
639 The documentation that comes along with your distribution will tell you how to do that.
642 * Darwin (MacOS/X) build environment::
643 * Cygwin (Windows) build environment::
644 * MinGW (Windows) build environment::
648 @c ==================================================================
649 @node Darwin (MacOS/X) build environment
650 @subsection Darwin (MacOS/X) build environment
652 In order to build tinc on Darwin, you need to install the MacOS/X Developer Tools
653 from @uref{http://developer.apple.com/tools/macosxtools.html} and
654 a recent version of Fink from @uref{http://fink.sourceforge.net/}.
656 After installation use fink to download and install the following packages:
657 autoconf25, automake, dlcompat, m4, openssl, zlib and lzo.
659 @c ==================================================================
660 @node Cygwin (Windows) build environment
661 @subsection Cygwin (Windows) build environment
663 If Cygwin hasn't already been installed, install it directly from
664 @uref{http://www.cygwin.com/}.
666 When tinc is compiled in a Cygwin environment, it can only be run in this environment,
667 but all programs, including those started outside the Cygwin environment, will be able to use the VPN.
668 It will also support all features.
670 @c ==================================================================
671 @node MinGW (Windows) build environment
672 @subsection MinGW (Windows) build environment
674 You will need to install the MinGW environment from @uref{http://www.mingw.org}.
676 When tinc is compiled using MinGW it runs natively under Windows,
677 it is not necessary to keep MinGW installed.
679 When detaching, tinc will install itself as a service,
680 which will be restarted automatically after reboots.
683 @c ==================================================================
685 @section System files
687 Before you can run tinc, you must make sure you have all the needed
688 files on your system.
696 @c ==================================================================
698 @subsection Device files
701 First, you'll need the special device file(s) that form the interface
702 between the kernel and the daemon.
704 The permissions for these files have to be such that only the super user
705 may read/write to this file. You'd want this, because otherwise
706 eavesdropping would become a bit too easy. This does, however, imply
707 that you'd have to run tincd as root.
709 If you use Linux and have a kernel version prior to 2.4.0, you have to make the
713 mknod -m 600 /dev/tap0 c 36 16
714 mknod -m 600 /dev/tap1 c 36 17
716 mknod -m 600 /dev/tap@emph{N} c 36 @emph{N+16}
719 There is a maximum of 16 ethertap devices.
721 If you use the universal tun/tap driver, you have to create the
722 following device file (unless it already exist):
725 mknod -m 600 /dev/tun c 10 200
728 If you use Linux, and you run the new 2.4 kernel using the devfs filesystem,
729 then the tun/tap device will probably be automatically generated as
732 Unlike the ethertap device, you do not need multiple device files if
733 you are planning to run multiple tinc daemons.
736 @c ==================================================================
738 @subsection Other files
740 @subsubheading @file{/etc/networks}
742 You may add a line to @file{/etc/networks} so that your VPN will get a
743 symbolic name. For example:
749 @subsubheading @file{/etc/services}
752 You may add this line to @file{/etc/services}. The effect is that you
753 may supply a @samp{tinc} as a valid port number to some programs. The
754 number 655 is registered with the IANA.
759 # Ivo Timmermans <ivo@@o2w.nl>
774 @c ==================================================================
776 @chapter Configuration
779 * Configuration introduction::
780 * Multiple networks::
781 * How connections work::
782 * Configuration files::
783 * Generating keypairs::
784 * Network interfaces::
785 * Example configuration::
788 @c ==================================================================
789 @node Configuration introduction
790 @section Configuration introduction
792 Before actually starting to configure tinc and editing files,
793 make sure you have read this entire section so you know what to expect.
794 Then, make it clear to yourself how you want to organize your VPN:
795 What are the nodes (computers running tinc)?
796 What IP addresses/subnets do they have?
797 What is the network mask of the entire VPN?
798 Do you need special firewall rules?
799 Do you have to set up masquerading or forwarding rules?
800 Do you want to run tinc in router mode or switch mode?
801 These questions can only be answered by yourself,
802 you will not find the answers in this documentation.
803 Make sure you have an adequate understanding of networks in general.
804 @cindex Network Administrators Guide
805 A good resource on networking is the
806 @uref{http://www.linuxdoc.org/LDP/nag2/, Linux Network Administrators Guide}.
808 If you have everything clearly pictured in your mind,
809 proceed in the following order:
810 First, generate the configuration files (@file{tinc.conf}, your host configuration file, @file{tinc-up} and perhaps @file{tinc-down}).
811 Then generate the keypairs.
812 Finally, distribute the host configuration files.
813 These steps are described in the subsections below.
816 @c ==================================================================
817 @node Multiple networks
818 @section Multiple networks
820 @cindex multiple networks
822 In order to allow you to run more than one tinc daemon on one computer,
823 for instance if your computer is part of more than one VPN,
824 you can assign a @var{netname} to your VPN.
825 It is not required if you only run one tinc daemon,
826 it doesn't even have to be the same on all the sites of your VPN,
827 but it is recommended that you choose one anyway.
829 We will asume you use a netname throughout this document.
830 This means that you call tincd with the -n argument,
831 which will assign a netname to this daemon.
833 The effect of this is that the daemon will set its configuration
834 root to @file{@value{sysconfdir}/tinc/@var{netname}/}, where @var{netname} is your argument to the -n
835 option. You'll notice that it appears in syslog as @file{tinc.@var{netname}}.
837 However, it is not strictly necessary that you call tinc with the -n
838 option. In this case, the network name would just be empty, and it will
839 be used as such. tinc now looks for files in @file{@value{sysconfdir}/tinc/}, instead of
840 @file{@value{sysconfdir}/tinc/@var{netname}/}; the configuration file should be @file{@value{sysconfdir}/tinc/tinc.conf},
841 and the host configuration files are now expected to be in @file{@value{sysconfdir}/tinc/hosts/}.
843 But it is highly recommended that you use this feature of tinc, because
844 it will be so much clearer whom your daemon talks to. Hence, we will
845 assume that you use it.
848 @c ==================================================================
849 @node How connections work
850 @section How connections work
852 When tinc starts up, it parses the command-line options and then
853 reads in the configuration file tinc.conf.
854 If it sees one or more `ConnectTo' values pointing to other tinc daemons in that file,
855 it will try to connect to those other daemons.
856 Whether this succeeds or not and whether `ConnectTo' is specified or not,
857 tinc will listen for incoming connection from other deamons.
858 If you did specify a `ConnectTo' value and the other side is not responding,
859 tinc will keep retrying.
860 This means that once started, tinc will stay running until you tell it to stop,
861 and failures to connect to other tinc daemons will not stop your tinc daemon
862 for trying again later.
863 This means you don't have to intervene if there are temporary network problems.
867 There is no real distinction between a server and a client in tinc.
868 If you wish, you can view a tinc daemon without a `ConnectTo' value as a server,
869 and one which does specify such a value as a client.
870 It does not matter if two tinc daemons have a `ConnectTo' value pointing to each other however.
873 @c ==================================================================
874 @node Configuration files
875 @section Configuration files
877 The actual configuration of the daemon is done in the file
878 @file{@value{sysconfdir}/tinc/@var{netname}/tinc.conf} and at least one other file in the directory
879 @file{@value{sysconfdir}/tinc/@var{netname}/hosts/}.
881 These file consists of comments (lines started with a #) or assignments
888 The variable names are case insensitive, and any spaces, tabs, newlines
889 and carriage returns are ignored. Note: it is not required that you put
890 in the `=' sign, but doing so improves readability. If you leave it
891 out, remember to replace it with at least one space character.
893 In this section all valid variables are listed in alphabetical order.
894 The default value is given between parentheses,
895 other comments are between square brackets.
898 * Main configuration variables::
899 * Host configuration variables::
905 @c ==================================================================
906 @node Main configuration variables
907 @subsection Main configuration variables
910 @cindex AddressFamily
911 @item AddressFamily = <ipv4|ipv6|any> (any)
912 This option affects the address family of listening and outgoing sockets.
913 If any is selected, then depending on the operating system
914 both IPv4 and IPv6 or just IPv6 listening sockets will be created.
916 @cindex BindToAddress
917 @item BindToAddress = <@var{address}> [experimental]
918 If your computer has more than one IPv4 or IPv6 address, tinc
919 will by default listen on all of them for incoming connections.
920 It is possible to bind only to a single address with this variable.
922 This option may not work on all platforms.
924 @cindex BindToInterface
925 @item BindToInterface = <@var{interface}> [experimental]
926 If you have more than one network interface in your computer, tinc will
927 by default listen on all of them for incoming connections. It is
928 possible to bind tinc to a single interface like eth0 or ppp0 with this
931 This option may not work on all platforms.
934 @item ConnectTo = <@var{name}>
935 Specifies which other tinc daemon to connect to on startup.
936 Multiple ConnectTo variables may be specified,
937 in which case outgoing connections to each specified tinc daemon are made.
938 The names should be known to this tinc daemon
939 (i.e., there should be a host configuration file for the name on the ConnectTo line).
941 If you don't specify a host with ConnectTo,
942 tinc won't try to connect to other daemons at all,
943 and will instead just listen for incoming connections.
946 @item Device = <@var{device}> (@file{/dev/tap0}, @file{/dev/net/tun} or other depending on platform)
947 The virtual network device to use.
948 Tinc will automatically detect what kind of device it is.
949 Note that you can only use one device per daemon.
950 Under Windows, use @var{Interface} instead of @var{Device}.
951 Note that you can only use one device per daemon.
952 See also @ref{Device files}.
955 @item Hostnames = <yes|no> (no)
956 This option selects whether IP addresses (both real and on the VPN)
957 should be resolved. Since DNS lookups are blocking, it might affect
958 tinc's efficiency, even stopping the daemon for a few seconds everytime
959 it does a lookup if your DNS server is not responding.
961 This does not affect resolving hostnames to IP addresses from the
965 @item Interface = <@var{interface}>
966 Defines the name of the interface corresponding to the virtual network device.
967 Depending on the operating system and the type of device this may or may not actually set the name of the interface.
968 Under Windows, this variable is used to select which network interface will be used.
969 If you specified a Device, this variable is almost always already correctly set.
972 @item Mode = <router|switch|hub> (router)
973 This option selects the way packets are routed to other daemons.
979 variables in the host configuration files will be used to form a routing table.
980 Only unicast packets of routable protocols (IPv4 and IPv6) are supported in this mode.
982 This is the default mode, and unless you really know you need another mode, don't change it.
986 In this mode the MAC addresses of the packets on the VPN will be used to
987 dynamically create a routing table just like an Ethernet switch does.
988 Unicast, multicast and broadcast packets of every protocol that runs over Ethernet are supported in this mode
989 at the cost of frequent broadcast ARP requests and routing table updates.
991 This mode is primarily useful if you want to bridge Ethernet segments.
995 This mode is almost the same as the switch mode, but instead
996 every packet will be broadcast to the other daemons
997 while no routing table is managed.
1001 @item KeyExpire = <@var{seconds}> (3600)
1002 This option controls the time the encryption keys used to encrypt the data
1003 are valid. It is common practice to change keys at regular intervals to
1004 make it even harder for crackers, even though it is thought to be nearly
1005 impossible to crack a single key.
1008 @item MACExpire = <@var{seconds}> (600)
1009 This option controls the amount of time MAC addresses are kept before they are removed.
1010 This only has effect when Mode is set to "switch".
1013 @item Name = <@var{name}> [required]
1014 This is a symbolic name for this connection. It can be anything
1017 @item PingTimeout = <@var{seconds}> (60)
1018 The number of seconds of inactivity that tinc will wait before sending a
1019 probe to the other end. If that other end doesn't answer within that
1020 same amount of seconds, the connection is terminated, and the others
1021 will be notified of this.
1023 @cindex PriorityInheritance
1024 @item PriorityInheritance = <yes|no> (no) [experimental]
1025 When this option is enabled the value of the TOS field of tunneled IPv4 packets
1026 will be inherited by the UDP packets that are sent out.
1029 @item PrivateKey = <@var{key}> [obsolete]
1030 This is the RSA private key for tinc. However, for safety reasons it is
1031 advised to store private keys of any kind in separate files. This prevents
1032 accidental eavesdropping if you are editting the configuration file.
1034 @cindex PrivateKeyFile
1035 @item PrivateKeyFile = <@var{path}> (@file{@value{sysconfdir}/tinc/@var{netname}/rsa_key.priv})
1036 This is the full path name of the RSA private key file that was
1037 generated by @samp{tincd --generate-keys}. It must be a full path, not a
1040 Note that there must be exactly one of PrivateKey
1042 specified in the configuration file.
1047 @c ==================================================================
1048 @node Host configuration variables
1049 @subsection Host configuration variables
1053 @item Address = <@var{IP address}|@var{hostname}> [recommended]
1054 This variable is only required if you want to connect to this host. It
1055 must resolve to the external IP address where the host can be reached,
1056 not the one that is internal to the VPN.
1059 @item Cipher = <@var{cipher}> (blowfish)
1060 The symmetric cipher algorithm used to encrypt UDP packets.
1061 Any cipher supported by OpenSSL is recognized.
1062 Furthermore, specifying "none" will turn off packet encryption.
1063 It is best to use only those ciphers which support CBC mode.
1066 @item Compression = <@var{level}> (0)
1067 This option sets the level of compression used for UDP packets.
1068 Possible values are 0 (off), 1 (fast zlib) and any integer up to 9 (best zlib),
1069 10 (fast lzo) and 11 (best lzo).
1072 @item Digest = <@var{digest}> (sha1)
1073 The digest algorithm used to authenticate UDP packets.
1074 Any digest supported by OpenSSL is recognized.
1075 Furthermore, specifying "none" will turn off packet authentication.
1077 @cindex IndirectData
1078 @item IndirectData = <yes|no> (no)
1079 This option specifies whether other tinc daemons besides the one you
1080 specified with ConnectTo can make a direct connection to you. This is
1081 especially useful if you are behind a firewall and it is impossible to
1082 make a connection from the outside to your tinc daemon. Otherwise, it
1083 is best to leave this option out or set it to no.
1086 @item MACLength = <@var{bytes}> (4)
1087 The length of the message authentication code used to authenticate UDP packets.
1088 Can be anything from 0
1089 up to the length of the digest produced by the digest algorithm.
1092 @item Port = <@var{port}> (655)
1093 This is the port this tinc daemon listens on.
1094 You can use decimal portnumbers or symbolic names (as listed in @file{/etc/services}).
1097 @item PublicKey = <@var{key}> [obsolete]
1098 This is the RSA public key for this host.
1100 @cindex PublicKeyFile
1101 @item PublicKeyFile = <@var{path}> [obsolete]
1102 This is the full path name of the RSA public key file that was generated
1103 by @samp{tincd --generate-keys}. It must be a full path, not a relative
1107 From version 1.0pre4 on tinc will store the public key directly into the
1108 host configuration file in PEM format, the above two options then are not
1109 necessary. Either the PEM format is used, or exactly
1110 @strong{one of the above two options} must be specified
1111 in each host configuration file, if you want to be able to establish a
1112 connection with that host.
1115 @item Subnet = <@var{address}[/@var{prefixlength}]>
1116 The subnet which this tinc daemon will serve.
1117 Tinc tries to look up which other daemon it should send a packet to by searching the appropiate subnet.
1118 If the packet matches a subnet,
1119 it will be sent to the daemon who has this subnet in his host configuration file.
1120 Multiple subnet lines can be specified for each daemon.
1122 Subnets can either be single MAC, IPv4 or IPv6 addresses,
1123 in which case a subnet consisting of only that single address is assumed,
1124 or they can be a IPv4 or IPv6 network address with a prefixlength.
1125 Shorthand notations are not supported.
1126 For example, IPv4 subnets must be in a form like 192.168.1.0/24,
1127 where 192.168.1.0 is the network address and 24 is the number of bits set in the netmask.
1128 Note that subnets like 192.168.1.1/24 are invalid!
1129 Read a networking HOWTO/FAQ/guide if you don't understand this.
1130 IPv6 subnets are notated like fec0:0:0:1:0:0:0:0/64.
1131 MAC addresses are notated like 0:1a:2b:3c:4d:5e.
1133 @cindex CIDR notation
1134 prefixlength is the number of bits set to 1 in the netmask part; for
1135 example: netmask 255.255.255.0 would become /24, 255.255.252.0 becomes
1136 /22. This conforms to standard CIDR notation as described in
1137 @uref{ftp://ftp.isi.edu/in-notes/rfc1519.txt, RFC1519}
1140 @item TCPonly = <yes|no> (no) [experimental]
1141 If this variable is set to yes, then the packets are tunnelled over a
1142 TCP connection instead of a UDP connection. This is especially useful
1143 for those who want to run a tinc daemon from behind a masquerading
1144 firewall, or if UDP packet routing is disabled somehow.
1145 Setting this options also implicitly sets IndirectData.
1149 @c ==================================================================
1154 Apart from reading the server and host configuration files,
1155 tinc can also run scripts at certain moments.
1156 Under Windows (not Cygwin), the scripts should have the extension .bat.
1160 @item @value{sysconfdir}/tinc/@var{netname}/tinc-up
1161 This is the most important script.
1162 If it is present it will be executed right after the tinc daemon has been
1163 started and has connected to the virtual network device.
1164 It should be used to set up the corresponding network interface,
1165 but can also be used to start other things.
1166 Under Windows you can use the Network Connections control panel instead of creating this script.
1169 @item @value{sysconfdir}/tinc/@var{netname}/tinc-down
1170 This script is started right before the tinc daemon quits.
1172 @item @value{sysconfdir}/tinc/@var{netname}/hosts/@var{host}-up
1173 This script is started when the tinc daemon with name @var{host} becomes reachable.
1175 @item @value{sysconfdir}/tinc/@var{netname}/hosts/@var{host}-down
1176 This script is started when the tinc daemon with name @var{host} becomes unreachable.
1179 @cindex environment variables
1180 The scripts are started without command line arguments,
1181 but can make use of certain environment variables.
1182 Under UNIX like operating systems the names of environment variables must be preceded by a $ in scripts.
1183 Under Windows, in @file{.bat} files, they have to be put between % signs.
1188 If a netname was specified, this environment variable contains it.
1192 Contains the name of this tinc daemon.
1196 Contains the name of the virtual network device that tinc uses.
1200 Contains the name of the virtual network interface that tinc uses.
1201 This should be used for commands like ifconfig.
1205 When a host becomes (un)reachable, this is set to its name.
1207 @cindex REMOTEADDRESS
1209 When a host becomes (un)reachable, this is set to its real address.
1213 When a host becomes (un)reachable,
1214 this is set to the port number it uses for communication with other tinc daemons.
1218 @c ==================================================================
1219 @node How to configure
1220 @subsection How to configure
1222 @subsubheading Step 1. Creating the main configuration file
1224 The main configuration file will be called @file{@value{sysconfdir}/tinc/@var{netname}/tinc.conf}.
1225 Adapt the following example to create a basic configuration file:
1228 Name = @var{yourname}
1229 Device = @file{/dev/tap0}
1232 Then, if you know to which other tinc daemon(s) yours is going to connect,
1233 add `ConnectTo' values.
1235 @subsubheading Step 2. Creating your host configuration file
1237 If you added a line containing `Name = yourname' in the main configuarion file,
1238 you will need to create a host configuration file @file{@value{sysconfdir}/tinc/@var{netname}/hosts/yourname}.
1239 Adapt the following example to create a host configuration file:
1242 Address = your.real.hostname.org
1243 Subnet = 192.168.1.0/24
1246 You can also use an IP address instead of a hostname.
1247 The `Subnet' specifies the address range that is local for @emph{your part of the VPN only}.
1248 If you have multiple address ranges you can specify more than one `Subnet'.
1249 You might also need to add a `Port' if you want your tinc daemon to run on a different port number than the default (655).
1252 @c ==================================================================
1253 @node Generating keypairs
1254 @section Generating keypairs
1256 @cindex key generation
1257 Now that you have already created the main configuration file and your host configuration file,
1258 you can easily create a public/private keypair by entering the following command:
1261 tincd -n @var{netname} -K
1264 Tinc will generate a public and a private key and ask you where to put them.
1265 Just press enter to accept the defaults.
1268 @c ==================================================================
1269 @node Network interfaces
1270 @section Network interfaces
1272 Before tinc can start transmitting data over the tunnel, it must
1273 set up the virtual network interface.
1275 First, decide which IP addresses you want to have associated with these
1276 devices, and what network mask they must have.
1278 Tinc will open a virtual network device (@file{/dev/tun}, @file{/dev/tap0} or similar),
1279 which will also create a network interface called something like @samp{tun0}, @samp{tap0}.
1280 If you are using the Linux tun/tap driver, the network interface will by default have the same name as the @var{netname}.
1281 Under Windows you can change the name of the network interface from the Network Connections control panel.
1284 You can configure the network interface by putting ordinary ifconfig, route, and other commands
1285 to a script named @file{@value{sysconfdir}/tinc/@var{netname}/tinc-up}.
1286 When tinc starts, this script will be executed. When tinc exits, it will execute the script named
1287 @file{@value{sysconfdir}/tinc/@var{netname}/tinc-down}, but normally you don't need to create that script.
1289 An example @file{tinc-up} script:
1293 ifconfig $INTERFACE 192.168.1.1 netmask 255.255.0.0
1296 This script gives the interface an IP address and a netmask.
1297 The kernel will also automatically add a route to this interface, so normally you don't need
1298 to add route commands to the @file{tinc-up} script.
1299 The kernel will also bring the interface up after this command.
1301 The netmask is the mask of the @emph{entire} VPN network, not just your
1304 The exact syntax of the ifconfig and route commands differs from platform to platform.
1305 You can look up the commands for setting addresses and adding routes in @ref{Platform specific information},
1306 but it is best to consult the manpages of those utilities on your platform.
1309 @c ==================================================================
1310 @node Example configuration
1311 @section Example configuration
1315 Imagine the following situation. Branch A of our example `company' wants to connect
1316 three branch offices in B, C and D using the Internet. All four offices
1317 have a 24/7 connection to the Internet.
1319 A is going to serve as the center of the network. B and C will connect
1320 to A, and D will connect to C. Each office will be assigned their own IP
1324 A: net 10.1.0.0 mask 255.255.0.0 gateway 10.1.54.1 internet IP 1.2.3.4
1325 B: net 10.2.0.0 mask 255.255.0.0 gateway 10.2.1.12 internet IP 2.3.4.5
1326 C: net 10.3.0.0 mask 255.255.0.0 gateway 10.3.69.254 internet IP 3.4.5.6
1327 D: net 10.4.0.0 mask 255.255.0.0 gateway 10.4.3.32 internet IP 4.5.6.7
1330 Here, ``gateway'' is the VPN IP address of the machine that is running the
1331 tincd, and ``internet IP'' is the IP address of the firewall, which does not
1332 need to run tincd, but it must do a port forwarding of TCP and UDP on port
1333 655 (unless otherwise configured).
1335 In this example, it is assumed that eth0 is the interface that points to
1336 the inner (physical) LAN of the office, although this could also be the
1337 same as the interface that leads to the Internet. The configuration of
1338 the real interface is also shown as a comment, to give you an idea of
1339 how these example host is set up. All branches use the netname `company'
1340 for this particular VPN.
1342 @subsubheading For Branch A
1344 @emph{BranchA} would be configured like this:
1346 In @file{@value{sysconfdir}/tinc/company/tinc-up}:
1349 # Real interface of internal network:
1350 # ifconfig eth0 10.1.54.1 netmask 255.255.0.0
1352 ifconfig $INTERFACE 10.1.54.1 netmask 255.0.0.0
1355 and in @file{@value{sysconfdir}/tinc/company/tinc.conf}:
1359 PrivateKeyFile = @value{sysconfdir}/tinc/company/rsa_key.priv
1363 On all hosts, @file{@value{sysconfdir}/tinc/company/hosts/BranchA} contains:
1366 Subnet = 10.1.0.0/16
1369 -----BEGIN RSA PUBLIC KEY-----
1371 -----END RSA PUBLIC KEY-----
1374 Note that the IP addresses of eth0 and tap0 are the same.
1375 This is quite possible, if you make sure that the netmasks of the interfaces are different.
1376 It is in fact recommended to give give both real internal network interfaces and tap interfaces the same IP address,
1377 since that will make things a lot easier to remember and set up.
1380 @subsubheading For Branch B
1382 In @file{@value{sysconfdir}/tinc/company/tinc-up}:
1385 # Real interface of internal network:
1386 # ifconfig eth0 10.2.43.8 netmask 255.255.0.0
1388 ifconfig $INTERFACE 10.2.1.12 netmask 255.0.0.0
1391 and in @file{@value{sysconfdir}/tinc/company/tinc.conf}:
1396 PrivateKeyFile = @value{sysconfdir}/tinc/company/rsa_key.priv
1399 Note here that the internal address (on eth0) doesn't have to be the
1400 same as on the tap0 device. Also, ConnectTo is given so that no-one can
1401 connect to this node.
1403 On all hosts, in @file{@value{sysconfdir}/tinc/company/hosts/BranchB}:
1406 Subnet = 10.2.0.0/16
1409 -----BEGIN RSA PUBLIC KEY-----
1411 -----END RSA PUBLIC KEY-----
1415 @subsubheading For Branch C
1417 In @file{@value{sysconfdir}/tinc/company/tinc-up}:
1420 # Real interface of internal network:
1421 # ifconfig eth0 10.3.69.254 netmask 255.255.0.0
1423 ifconfig $INTERFACE 10.3.69.254 netmask 255.0.0.0
1426 and in @file{@value{sysconfdir}/tinc/company/tinc.conf}:
1434 C already has another daemon that runs on port 655, so they have to
1435 reserve another port for tinc. It knows the portnumber it has to listen on
1436 from it's own host configuration file.
1438 On all hosts, in @file{@value{sysconfdir}/tinc/company/hosts/BranchC}:
1442 Subnet = 10.3.0.0/16
1445 -----BEGIN RSA PUBLIC KEY-----
1447 -----END RSA PUBLIC KEY-----
1451 @subsubheading For Branch D
1453 In @file{@value{sysconfdir}/tinc/company/tinc-up}:
1456 # Real interface of internal network:
1457 # ifconfig eth0 10.4.3.32 netmask 255.255.0.0
1459 ifconfig $INTERFACE 10.4.3.32 netmask 255.0.0.0
1462 and in @file{@value{sysconfdir}/tinc/company/tinc.conf}:
1467 Device = /dev/net/tun
1468 PrivateKeyFile = @value{sysconfdir}/tinc/company/rsa_key.priv
1471 D will be connecting to C, which has a tincd running for this network on
1472 port 2000. It knows the port number from the host configuration file.
1473 Also note that since D uses the tun/tap driver, the network interface
1474 will not be called `tun' or `tap0' or something like that, but will
1475 have the same name as netname.
1477 On all hosts, in @file{@value{sysconfdir}/tinc/company/hosts/BranchD}:
1480 Subnet = 10.4.0.0/16
1483 -----BEGIN RSA PUBLIC KEY-----
1485 -----END RSA PUBLIC KEY-----
1488 @subsubheading Key files
1490 A, B, C and D all have generated a public/private keypair with the following command:
1496 The private key is stored in @file{@value{sysconfdir}/tinc/company/rsa_key.priv},
1497 the public key is put into the host configuration file in the @file{@value{sysconfdir}/tinc/company/hosts/} directory.
1498 During key generation, tinc automatically guesses the right filenames based on the -n option and
1499 the Name directive in the @file{tinc.conf} file (if it is available).
1501 @subsubheading Starting
1503 After each branch has finished configuration and they have distributed
1504 the host configuration files amongst them, they can start their tinc daemons.
1505 They don't necessarily have to wait for the other branches to have started
1506 their daemons, tinc will try connecting until they are available.
1509 @c ==================================================================
1511 @chapter Running tinc
1513 If everything else is done, you can start tinc by typing the following command:
1516 tincd -n @var{netname}
1520 Tinc will detach from the terminal and continue to run in the background like a good daemon.
1521 If there are any problems however you can try to increase the debug level
1522 and look in the syslog to find out what the problems are.
1526 * Solving problems::
1531 @c ==================================================================
1532 @node Runtime options
1533 @section Runtime options
1535 Besides the settings in the configuration file, tinc also accepts some
1536 command line options.
1538 @cindex command line
1539 @cindex runtime options
1543 @item -c, --config=@var{path}
1544 Read configuration options from the directory @var{path}. The default is
1545 @file{@value{sysconfdir}/tinc/@var{netname}/}.
1547 @item -D, --no-detach
1548 Don't fork and detach.
1549 This will also disable the automatic restart mechanism for fatal errors.
1552 @item -d, --debug=@var{level}
1553 Set debug level to @var{level}. The higher the debug level, the more gets
1554 logged. Everything goes via syslog.
1556 @item -k, --kill[=@var{signal}]
1557 Attempt to kill a running tincd (optionally with the specified @var{signal} instead of SIGTERM) and exit.
1558 Use it in conjunction with the -n option to make sure you kill the right tinc daemon.
1559 Under native Windows the optional argument is ignored,
1560 the service will always be stopped and removed.
1562 @item -n, --net=@var{netname}
1563 Use configuration for net @var{netname}. @xref{Multiple networks}.
1565 @item -K, --generate-keys[=@var{bits}]
1566 Generate public/private keypair of @var{bits} length. If @var{bits} is not specified,
1567 1024 is the default. tinc will ask where you want to store the files,
1568 but will default to the configuration directory (you can use the -c or -n option
1569 in combination with -K). After that, tinc will quit.
1572 Lock tinc into main memory.
1573 This will prevent sensitive data like shared private keys to be written to the system swap files/partitions.
1575 @item --logfile[=@var{file}]
1576 Write log entries to a file instead of to the system logging facility.
1577 If @var{file} is omitted, the default is @file{@value{localstatedir}/log/tinc.@var{netname}.log}.
1579 @item --pidfile=@var{file}
1580 Write PID to @var{file} instead of @file{@value{localstatedir}/run/tinc.@var{netname}.pid}.
1582 @item --bypass-security
1583 Disables encryption and authentication.
1584 Only useful for debugging.
1587 Display a short reminder of these runtime options and terminate.
1590 Output version information and exit.
1594 @c ==================================================================
1595 @node Solving problems
1596 @section Solving problems
1598 If tinc starts without problems, but if the VPN doesn't work, you will have to find the cause of the problem.
1599 The first thing to do is to start tinc with a high debug level in the foreground,
1600 so you can directly see everything tinc logs:
1603 tincd -n @var{netname} -d5 -D
1606 If tinc does not log any error messages, then you might want to check the following things:
1609 @item @file{tinc-up} script
1610 Does this script contain the right commands?
1611 Normally you must give the interface the address of this host on the VPN, and the netmask must be big enough so that the entire VPN is covered.
1614 Does the Subnet (or Subnets) in the host configuration file of this host match the portion of the VPN that belongs to this host?
1616 @item Firewalls and NATs
1617 Do you have a firewall or a NAT device (a masquerading firewall or perhaps an ADSL router that performs masquerading)?
1618 If so, check that it allows TCP and UDP traffic on port 655.
1619 If it masquerades and the host running tinc is behind it, make sure that it forwards TCP and UDP traffic to port 655 to the host running tinc.
1620 You can add @samp{TCPOnly = yes} to your host config file to force tinc to only use a single TCP connection,
1621 this works through most firewalls and NATs.
1626 @c ==================================================================
1627 @node Error messages
1628 @section Error messages
1630 What follows is a list of the most common error messages you might find in the logs.
1631 Some of them will only be visible if the debug level is high enough.
1634 @item Could not open /dev/tap0: No such device
1637 @item You forgot to `modprobe netlink_dev' or `modprobe ethertap'.
1638 @item You forgot to compile `Netlink device emulation' in the kernel.
1641 @item Can't write to /dev/net/tun: No such device
1644 @item You forgot to `modprobe tun'.
1645 @item You forgot to compile `Universal TUN/TAP driver' in the kernel.
1646 @item The tun device is located somewhere else in @file{/dev/}.
1649 @item Network address and prefix length do not match!
1652 @item The Subnet field must contain a @emph{network} address, trailing bits should be 0.
1653 @item If you only want to use one IP address, set the netmask to /32.
1656 @item Error reading RSA key file `rsa_key.priv': No such file or directory
1659 @item You forgot to create a public/private keypair.
1660 @item Specify the complete pathname to the private key file with the @samp{PrivateKeyFile} option.
1663 @item Warning: insecure file permissions for RSA private key file `rsa_key.priv'!
1666 @item The private key file is readable by users other than root.
1667 Use chmod to correct the file permissions.
1670 @item Creating metasocket failed: Address family not supported
1673 @item By default tinc tries to create both IPv4 and IPv6 sockets.
1674 On some platforms this might not be implemented.
1675 If the logs show @samp{Ready} later on, then at least one metasocket was created,
1676 and you can ignore this message.
1677 You can add @samp{AddressFamily = ipv4} to @file{tinc.conf} to prevent this from happening.
1680 @item Cannot route packet: unknown IPv4 destination 1.2.3.4
1683 @item You try to send traffic to a host on the VPN for which no Subnet is known.
1684 @item If it is a broadcast address (ending in .255), it probably is a samba server or a Windows host sending broadcast packets.
1688 @item Cannot route packet: ARP request for unknown address 1.2.3.4
1691 @item You try to send traffic to a host on the VPN for which no Subnet is known.
1694 @item Packet with destination 1.2.3.4 is looping back to us!
1697 @item Something is not configured right. Packets are being sent out to the
1698 virtual network device, but according to the Subnet directives in your host configuration
1699 file, those packets should go to your own host. Most common mistake is that
1700 you have a Subnet line in your host configuration file with a prefix length which is
1701 just as large as the prefix of the virtual network interface. The latter should in almost all
1702 cases be larger. Rethink your configuration.
1703 Note that you will only see this message if you specified a debug
1704 level of 5 or higher!
1705 @item Chances are that a @samp{Subnet = ...} line in the host configuration file of this tinc daemon is wrong.
1706 Change it to a subnet that is accepted locally by another interface,
1707 or if that is not the case, try changing the prefix length into /32.
1710 @item Node foo (1.2.3.4) is not reachable
1713 @item Node foo does not have a connection anymore, its tinc daemon is not running or its connection to the Internet is broken.
1716 @item Received UDP packet from unknown source 1.2.3.4 (port 12345)
1719 @item If you see this only sporadically, it is harmless and caused by a node sending packets using an old key.
1720 @item If you see this often and another node is not reachable anymore, then a NAT (masquerading firewall) is changing the source address of UDP packets.
1721 You can add @samp{TCPOnly = yes} to host configuration files to force all VPN traffic to go over a TCP connection.
1724 @item Got bad/bogus/unauthorized REQUEST from foo (1.2.3.4 port 12345)
1727 @item Node foo does not have the right public/private keypair.
1728 Generate new keypairs and distribute them again.
1729 @item An attacker tries to gain access to your VPN.
1730 @item A network error caused corruption of metadata sent from foo.
1735 @c ==================================================================
1736 @node Sending bug reports
1737 @section Sending bug reports
1739 If you really can't find the cause of a problem, or if you suspect tinc is not working right,
1740 you can send us a bugreport, see @ref{Contact information}.
1741 Be sure to include the following information in your bugreport:
1744 @item A clear description of what you are trying to achieve and what the problem is.
1745 @item What platform (operating system, version, hardware architecture) and which version of tinc you use.
1746 @item If compiling tinc fails, a copy of @file{config.log} and the error messages you get.
1747 @item Otherwise, a copy of @file{tinc.conf}, @file{tinc-up} and all files in the @file{hosts/} directory.
1748 @item The output of the commands @samp{ifconfig -a} and @samp{route -n} (or @samp{netstat -rn} if that doesn't work).
1749 @item The output of any command that fails to work as it should (like ping or traceroute).
1752 @c ==================================================================
1753 @node Technical information
1754 @chapter Technical information
1759 * The meta-protocol::
1764 @c ==================================================================
1765 @node The connection
1766 @section The connection
1769 Tinc is a daemon that takes VPN data and transmit that to another host
1770 computer over the existing Internet infrastructure.
1774 * The meta-connection::
1778 @c ==================================================================
1779 @node The UDP tunnel
1780 @subsection The UDP tunnel
1782 @cindex virtual network device
1784 The data itself is read from a character device file, the so-called
1785 @emph{virtual network device}. This device is associated with a network
1786 interface. Any data sent to this interface can be read from the device,
1787 and any data written to the device gets sent from the interface.
1788 There are two possible types of virtual network devices:
1789 `tun' style, which are point-to-point devices which can only handle IPv4 and/or IPv6 packets,
1790 and `tap' style, which are Ethernet devices and handle complete Ethernet frames.
1792 So when tinc reads an Ethernet frame from the device, it determines its
1793 type. When tinc is in it's default routing mode, it can handle IPv4 and IPv6
1794 packets. Depending on the Subnet lines, it will send the packets off to their destination IP address.
1795 In the `switch' and `hub' mode, tinc will use broadcasts and MAC address discovery
1796 to deduce the destination of the packets.
1797 Since the latter modes only depend on the link layer information,
1798 any protocol that runs over Ethernet is supported (for instance IPX and Appletalk).
1799 However, only `tap' style devices provide this information.
1801 After the destination has been determined,
1802 the packet will be compressed (optionally),
1803 a sequence number will be added to the packet,
1804 the packet will then be encrypted
1805 and a message authentication code will be appended.
1807 @cindex encapsulating
1809 When that is done, time has come to actually transport the
1810 packet to the destination computer. We do this by sending the packet
1811 over an UDP connection to the destination host. This is called
1812 @emph{encapsulating}, the VPN packet (though now encrypted) is
1813 encapsulated in another IP datagram.
1815 When the destination receives this packet, the same thing happens, only
1816 in reverse. So it checks the message authentication code, decrypts the contents of the UDP datagram,
1817 checks the sequence number
1818 and writes the decrypted information to its own virtual network device.
1820 If the virtual network device is a `tun' device (a point-to-point tunnel),
1821 there is no problem for the kernel to accept a packet.
1822 However, if it is a `tap' device (this is the only available type on FreeBSD),
1823 the destination MAC address must match that of the virtual network interface.
1824 If tinc is in it's default routing mode, ARP does not work, so the correct destination MAC
1825 can not be known by the sending host.
1826 Tinc solves this by letting the receiving end detect the MAC address of its own virtual network interface
1827 and overwriting the destination MAC address of the received packet.
1829 In switch or hub modes ARP does work so the sender already knows the correct destination MAC address.
1830 In those modes every interface should have a unique MAC address, so make sure they are not the same.
1831 Because switch and hub modes rely on MAC addresses to function correctly,
1832 these modes cannot be used on the following operating systems which don't have a `tap' style virtual network device:
1833 OpenBSD, NetBSD, Darwin and Solaris.
1836 @c ==================================================================
1837 @node The meta-connection
1838 @subsection The meta-connection
1840 Having only a UDP connection available is not enough. Though suitable
1841 for transmitting data, we want to be able to reliably send other
1842 information, such as routing and session key information to somebody.
1845 TCP is a better alternative, because it already contains protection
1846 against information being lost, unlike UDP.
1848 So we establish two connections. One for the encrypted VPN data, and one
1849 for other information, the meta-data. Hence, we call the second
1850 connection the meta-connection. We can now be sure that the
1851 meta-information doesn't get lost on the way to another computer.
1853 @cindex data-protocol
1854 @cindex meta-protocol
1855 Like with any communication, we must have a protocol, so that everybody
1856 knows what everything stands for, and how she should react. Because we
1857 have two connections, we also have two protocols. The protocol used for
1858 the UDP data is the ``data-protocol,'' the other one is the
1861 The reason we don't use TCP for both protocols is that UDP is much
1862 better for encapsulation, even while it is less reliable. The real
1863 problem is that when TCP would be used to encapsulate a TCP stream
1864 that's on the private network, for every packet sent there would be
1865 three ACKs sent instead of just one. Furthermore, if there would be
1866 a timeout, both TCP streams would sense the timeout, and both would
1867 start re-sending packets.
1870 @c ==================================================================
1871 @node The meta-protocol
1872 @section The meta-protocol
1874 The meta protocol is used to tie all tinc daemons together, and
1875 exchange information about which tinc daemon serves which virtual
1878 The meta protocol consists of requests that can be sent to the other
1879 side. Each request has a unique number and several parameters. All
1880 requests are represented in the standard ASCII character set. It is
1881 possible to use tools such as telnet or netcat to connect to a tinc
1882 daemon started with the --bypass-security option
1883 and to read and write requests by hand, provided that one
1884 understands the numeric codes sent.
1886 The authentication scheme is described in @ref{Authentication protocol}. After a
1887 successful authentication, the server and the client will exchange all the
1888 information about other tinc daemons and subnets they know of, so that both
1889 sides (and all the other tinc daemons behind them) have their information
1896 --------------------------------------------------------------------------
1897 origin ADD_EDGE node1 node2 21.32.43.54 655 222 0
1898 | | | | | +-> options
1899 | | | | +----> weight
1900 | | | +--------> UDP port of node2
1901 | | +----------------> real address of node2
1902 | +-------------------------> name of destination node
1903 +-------------------------------> name of source node
1905 origin ADD_SUBNET node 192.168.1.0/24
1906 | | +--> prefixlength
1907 | +--------> network address
1908 +------------------> owner of this subnet
1909 --------------------------------------------------------------------------
1912 The ADD_EDGE messages are to inform other tinc daemons that a connection between
1913 two nodes exist. The address of the destination node is available so that
1914 VPN packets can be sent directly to that node.
1916 The ADD_SUBNET messages inform other tinc daemons that certain subnets belong
1917 to certain nodes. tinc will use it to determine to which node a VPN packet has
1924 ------------------------------------------------------------------
1925 DEL_EDGE node1 node2
1926 | +----> name of destination node
1927 +----------> name of source node
1929 DEL_SUBNET node 192.168.1.0/24
1930 | | +--> prefixlength
1931 | +--------> network address
1932 +------------------> owner of this subnet
1933 ------------------------------------------------------------------
1936 In case a connection between two daemons is closed or broken, DEL_EDGE messages
1937 are sent to inform the other daemons of that fact. Each daemon will calculate a
1938 new route to the the daemons, or mark them unreachable if there isn't any.
1945 ------------------------------------------------------------------
1946 REQ_KEY origin destination
1947 | +--> name of the tinc daemon it wants the key from
1948 +----------> name of the daemon that wants the key
1950 ANS_KEY origin destination 4ae0b0a82d6e0078 91 64 4
1951 | | \______________/ | | +--> MAC length
1952 | | | | +-----> digest algorithm
1953 | | | +--------> cipher algorithm
1954 | | +--> 128 bits key
1955 | +--> name of the daemon that wants the key
1956 +----------> name of the daemon that uses this key
1959 +--> daemon that has changed it's packet key
1960 --------------------------------------------------------------------------
1963 The keys used to encrypt VPN packets are not sent out directly. This is
1964 because it would generate a lot of traffic on VPNs with many daemons, and
1965 chances are that not every tinc daemon will ever send a packet to every
1966 other daemon. Instead, if a daemon needs a key it sends a request for it
1967 via the meta connection of the nearest hop in the direction of the
1974 --------------------------------------------------------------------------
1977 --------------------------------------------------------------------------
1980 There is also a mechanism to check if hosts are still alive. Since network
1981 failures or a crash can cause a daemon to be killed without properly
1982 shutting down the TCP connection, this is necessary to keep an up to date
1983 connection list. PINGs are sent at regular intervals, except when there
1984 is also some other traffic. A little bit of salt (random data) is added
1985 with each PING and PONG message, to make sure that long sequences of PING/PONG
1986 messages without any other traffic won't result in known plaintext.
1988 This basically covers what is sent over the meta connection by tinc.
1991 @c ==================================================================
1997 Tinc got its name from ``TINC,'' short for @emph{There Is No Cabal}; the
1998 alleged Cabal was/is an organisation that was said to keep an eye on the
1999 entire Internet. As this is exactly what you @emph{don't} want, we named
2000 the tinc project after TINC.
2003 But in order to be ``immune'' to eavesdropping, you'll have to encrypt
2004 your data. Because tinc is a @emph{Secure} VPN (SVPN) daemon, it does
2005 exactly that: encrypt.
2006 Tinc by default uses blowfish encryption with 128 bit keys in CBC mode, 32 bit
2007 sequence numbers and 4 byte long message authentication codes to make sure
2008 eavesdroppers cannot get and cannot change any information at all from the
2009 packets they can intercept. The encryption algorithm and message authentication
2010 algorithm can be changed in the configuration. The length of the message
2011 authentication codes is also adjustable. The length of the key for the
2012 encryption algorithm is always the default length used by OpenSSL.
2015 * Authentication protocol::
2016 * Encryption of network packets::
2021 @c ==================================================================
2022 @node Authentication protocol
2023 @subsection Authentication protocol
2025 @cindex authentication
2026 A new scheme for authentication in tinc has been devised, which offers some
2027 improvements over the protocol used in 1.0pre2 and 1.0pre3. Explanation is
2037 --------------------------------------------------------------------------
2038 client <attempts connection>
2040 server <accepts connection>
2044 +-------> name of tinc daemon
2048 +-------> name of tinc daemon
2050 client META_KEY 5f0823a93e35b69e...7086ec7866ce582b
2051 \_________________________________/
2052 +-> RSAKEYLEN bits totally random string S1,
2053 encrypted with server's public RSA key
2055 server META_KEY 6ab9c1640388f8f0...45d1a07f8a672630
2056 \_________________________________/
2057 +-> RSAKEYLEN bits totally random string S2,
2058 encrypted with client's public RSA key
2061 - the client will symmetrically encrypt outgoing traffic using S1
2062 - the server will symmetrically encrypt outgoing traffic using S2
2064 client CHALLENGE da02add1817c1920989ba6ae2a49cecbda0
2065 \_________________________________/
2066 +-> CHALLEN bits totally random string H1
2068 server CHALLENGE 57fb4b2ccd70d6bb35a64c142f47e61d57f
2069 \_________________________________/
2070 +-> CHALLEN bits totally random string H2
2072 client CHAL_REPLY 816a86
2073 +-> 160 bits SHA1 of H2
2075 server CHAL_REPLY 928ffe
2076 +-> 160 bits SHA1 of H1
2078 After the correct challenge replies are received, both ends have proved
2079 their identity. Further information is exchanged.
2081 client ACK 655 123 0
2083 | +----> estimated weight
2084 +--------> listening port of client
2086 server ACK 655 321 0
2088 | +----> estimated weight
2089 +--------> listening port of server
2090 --------------------------------------------------------------------------
2093 This new scheme has several improvements, both in efficiency and security.
2095 First of all, the server sends exactly the same kind of messages over the wire
2096 as the client. The previous versions of tinc first authenticated the client,
2097 and then the server. This scheme even allows both sides to send their messages
2098 simultaneously, there is no need to wait for the other to send something first.
2099 This means that any calculations that need to be done upon sending or receiving
2100 a message can also be done in parallel. This is especially important when doing
2101 RSA encryption/decryption. Given that these calculations are the main part of
2102 the CPU time spent for the authentication, speed is improved by a factor 2.
2104 Second, only one RSA encrypted message is sent instead of two. This reduces the
2105 amount of information attackers can see (and thus use for a cryptographic
2106 attack). It also improves speed by a factor two, making the total speedup a
2109 Third, and most important:
2110 The symmetric cipher keys are exchanged first, the challenge is done
2111 afterwards. In the previous authentication scheme, because a man-in-the-middle
2112 could pass the challenge/chal_reply phase (by just copying the messages between
2113 the two real tinc daemons), but no information was exchanged that was really
2114 needed to read the rest of the messages, the challenge/chal_reply phase was of
2115 no real use. The man-in-the-middle was only stopped by the fact that only after
2116 the ACK messages were encrypted with the symmetric cipher. Potentially, it
2117 could even send it's own symmetric key to the server (if it knew the server's
2118 public key) and read some of the metadata the server would send it (it was
2119 impossible for the mitm to read actual network packets though). The new scheme
2120 however prevents this.
2122 This new scheme makes sure that first of all, symmetric keys are exchanged. The
2123 rest of the messages are then encrypted with the symmetric cipher. Then, each
2124 side can only read received messages if they have their private key. The
2125 challenge is there to let the other side know that the private key is really
2126 known, because a challenge reply can only be sent back if the challenge is
2127 decrypted correctly, and that can only be done with knowledge of the private
2130 Fourth: the first thing that is sent via the symmetric cipher encrypted
2131 connection is a totally random string, so that there is no known plaintext (for
2132 an attacker) in the beginning of the encrypted stream.
2135 @c ==================================================================
2136 @node Encryption of network packets
2137 @subsection Encryption of network packet
2140 A data packet can only be sent if the encryption key is known to both
2141 parties, and the connection is activated. If the encryption key is not
2142 known, a request is sent to the destination using the meta connection
2143 to retrieve it. The packet is stored in a queue while waiting for the
2147 The UDP packet containing the network packet from the VPN has the following layout:
2150 ... | IP header | UDP header | seqno | VPN packet | MAC | UDP trailer
2151 \___________________/\_____/
2153 V +---> digest algorithm
2154 Encrypted with symmetric cipher
2157 So, the entire VPN packet is encrypted using a symmetric cipher, including a 32 bits
2158 sequence number that is added in front of the actual VPN packet, to act as a unique
2159 IV for each packet and to prevent replay attacks. A message authentication code
2160 is added to the UDP packet to prevent alteration of packets. By default the
2161 first 4 bytes of the digest are used for this, but this can be changed using
2162 the MACLength configuration variable.
2164 @c ==================================================================
2165 @node Security issues
2166 @section Security issues
2168 In August 2000, we discovered the existence of a security hole in all versions
2169 of tinc up to and including 1.0pre2. This had to do with the way we exchanged
2170 keys. Since then, we have been working on a new authentication scheme to make
2171 tinc as secure as possible. The current version uses the OpenSSL library and
2172 uses strong authentication with RSA keys.
2174 On the 29th of December 2001, Jerome Etienne posted a security analysis of tinc
2175 1.0pre4. Due to a lack of sequence numbers and a message authentication code
2176 for each packet, an attacker could possibly disrupt certain network services or
2177 launch a denial of service attack by replaying intercepted packets. The current
2178 version adds sequence numbers and message authentication codes to prevent such
2181 On the 15th of September 2003, Peter Gutmann posted a security analysis of tinc
2182 1.0.1. He argues that the 32 bit sequence number used by tinc is not a good IV,
2183 that tinc's default length of 4 bytes for the MAC is too short, and he doesn't
2184 like tinc's use of RSA during authentication. We do not know of a security hole
2185 in this version of tinc, but tinc's security is not as strong as TLS or IPsec.
2186 We will address these issues in tinc 2.0.
2188 Cryptography is a hard thing to get right. We cannot make any
2189 guarantees. Time, review and feedback are the only things that can
2190 prove the security of any cryptographic product. If you wish to review
2191 tinc or give us feedback, you are stronly encouraged to do so.
2194 @c ==================================================================
2195 @node Platform specific information
2196 @chapter Platform specific information
2199 * Interface configuration::
2203 @c ==================================================================
2204 @node Interface configuration
2205 @section Interface configuration
2207 When configuring an interface, one normally assigns it an address and a
2208 netmask. The address uniquely identifies the host on the network attached to
2209 the interface. The netmask, combined with the address, forms a subnet. It is
2210 used to add a route to the routing table instructing the kernel to send all
2211 packets which fall into that subnet to that interface. Because all packets for
2212 the entire VPN should go to the virtual network interface used by tinc, the
2213 netmask should be such that it encompasses the entire VPN.
2217 @multitable {Darwin (MacOS/X)} {ifconfig route add -bla network address netmask netmask prefixlength interface}
2219 @tab @code{ifconfig} @var{interface} @var{address} @code{netmask} @var{netmask}
2220 @item Linux iproute2
2221 @tab @code{ip addr add} @var{address}@code{/}@var{prefixlength} @code{dev} @var{interface}
2223 @tab @code{ifconfig} @var{interface} @var{address} @code{netmask} @var{netmask}
2225 @tab @code{ifconfig} @var{interface} @var{address} @code{netmask} @var{netmask}
2227 @tab @code{ifconfig} @var{interface} @var{address} @code{netmask} @var{netmask}
2229 @tab @code{ifconfig} @var{interface} @var{address} @code{netmask} @var{netmask}
2230 @item Darwin (MacOS/X)
2231 @tab @code{ifconfig} @var{interface} @var{address} @code{netmask} @var{netmask}
2233 @tab @code{netsh interface ip set address} @var{interface} @code{static} @var{address} @var{netmask}
2239 @multitable {Darwin (MacOS/X)} {ifconfig route add -bla network address netmask netmask prefixlength interface}
2241 @tab @code{ifconfig} @var{interface} @code{add} @var{address}@code{/}@var{prefixlength}
2243 @tab @code{ifconfig} @var{interface} @code{inet6} @var{address} @code{prefixlen} @var{prefixlength}
2245 @tab @code{ifconfig} @var{interface} @code{inet6} @var{address} @code{prefixlen} @var{prefixlength}
2247 @tab @code{ifconfig} @var{interface} @code{inet6} @var{address} @code{prefixlen} @var{prefixlength}
2249 @tab @code{ifconfig} @var{interface} @code{inet6 addif} @var{address}@code{/}@var{prefixlength}
2250 @item Darwin (MacOS/X)
2251 @tab @code{ifconfig} @var{interface} @code{inet6} @var{address} @code{prefixlen} @var{prefixlength}
2253 @tab @code{netsh interface ipv6 add address} @var{interface} @code{static} @var{address}/@var{prefixlength}
2257 @c ==================================================================
2261 In some cases it might be necessary to add more routes to the virtual network
2262 interface. There are two ways to indicate which interface a packet should go
2263 to, one is to use the name of the interface itself, another way is to specify
2264 the (local) address that is assigned to that interface (@var{local_address}). The
2265 former way is unambiguous and therefore preferable, but not all platforms
2268 Adding routes to IPv4 subnets:
2270 @multitable {Darwin (MacOS/X)} {ifconfig route add -bla network address netmask netmask prefixlength interface}
2272 @tab @code{route add -net} @var{network_address} @code{netmask} @var{netmask} @var{interface}
2273 @item Linux iproute2
2274 @tab @code{ip route add} @var{network_address}@code{/}@var{prefixlength} @code{dev} @var{interface}
2276 @tab @code{route add} @var{network_address}@code{/}@var{prefixlength} @var{local_address}
2278 @tab @code{route add} @var{network_address}@code{/}@var{prefixlength} @var{local_address}
2280 @tab @code{route add} @var{network_address}@code{/}@var{prefixlength} @var{local_address}
2282 @item Darwin (MacOS/X)
2283 @tab @code{route add} @var{network_address}@code{/}@var{prefixlength} @var{local_address}
2287 Adding routes to IPv6 subnets:
2289 @multitable {Darwin (MacOS/X)} {ifconfig route add -bla network address netmask netmask prefixlength interface}
2291 @tab @code{route add -A inet6} @var{network_address}@code{/}@var{prefixlength} @var{interface}
2292 @item Linux iproute2
2293 @tab @code{ip route add} @var{network_address}@code{/}@var{prefixlength} @code{dev} @var{interface}
2297 @item Darwin (MacOS/X)
2299 @tab @code{netsh interface ipv6 add route} @var{network address}/@var{prefixlength} @var{interface}
2303 @c ==================================================================
2309 * Contact Information::
2314 @c ==================================================================
2315 @node Contact Information
2316 @section Contact information
2319 Tinc's website is at @url{http://tinc.nl.linux.org/},
2320 this server is located in the Netherlands.
2323 We have an IRC channel on the FreeNode and OFTC IRC networks. Connect to
2324 @uref{http://www.freenode.net/, irc.freenode.net}
2326 @uref{http://www.oftc.net/, irc.oftc.net}
2327 and join channel #tinc.
2330 @c ==================================================================
2335 @item Ivo Timmermans (zarq) (@email{ivo@@o2w.nl})
2336 @item Guus Sliepen (guus) (@email{guus@@sliepen.eu.org})
2339 We have received a lot of valuable input from users. With their help,
2340 tinc has become the flexible and robust tool that it is today. We have
2341 composed a list of contributions, in the file called @file{THANKS} in
2342 the source distribution.
2345 @c ==================================================================
2347 @unnumbered Concept Index
2349 @c ==================================================================
2353 @c ==================================================================