1 secnet - flexible VPN software
6 Copyright 1995-2003 Stephen Early <steve@greenend.org.uk>
7 Copyright 2002-2014 Ian Jackson <ijackson@chiark.greenend.org.uk>
8 Copyright 1991 Massachusetts Institute of Technology
9 Copyright 1998 Ross Anderson, Eli Biham, Lars Knudsen
10 Copyright 1993 Colin Plumb
11 Copyright 1998 James H. Brown, Steve Reid
12 Copyright 2000 Vincent Rijmen, Antoon Bosselaers, Paulo Barreto
13 Copyright 2001 Saul Kravitz
14 Copyright 2004 Fabrice Bellard
15 Copyright 2002 Guido Draheim
16 Copyright 2005-2010 Free Software Foundation, Inc.
17 Copyright 1995-2001 Jonathan Amery
18 Copyright 1995-2003 Peter Benie
19 Copyright 2011 Richard Kettlewell
20 Copyright 2012 Matthew Vernon
21 Copyright 2013 Mark Wooding
22 Copyright 1995-2013 Simon Tatham
24 secnet is distributed under the terms of the GNU General Public
25 License, version 3 or later. Some individual files have more
26 permissive licences; where this is the case, it is documented in the
27 header comment for the files in question.
29 secnet is distributed in the hope that it will be useful, but WITHOUT
30 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
31 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
34 The file COPYING contains a copy of the GNU GPL v3.
39 secnet allows large virtual private networks to be constructed
40 spanning multiple separate sites. It is designed for the case where a
41 private network connecting many hosts is 'hidden' behind a single
42 globally-routable IP address, but can also be applied in other
43 circumstances. It communicates entirely using UDP, and works well
44 with gateways that implement network address translation.
46 If you are installing secnet to join an existing VPN, you should read
47 the 'INSTALL' file and your particular VPN's documentation now. You
48 may need to refer back to this file for information on the netlink and
49 comm sections of the configuration file.
51 If you are thinking about setting up a new VPN of any size (from one
52 providing complete links between multiple sites to a simple
53 laptop-to-host link), read the section in this file on 'Creating a
56 * Mailing lists and bug reporting
58 There are two mailing lists associated with secnet: an 'announce' list
59 and a 'discuss' list. Their addresses are:
60 http://www.chiark.greenend.org.uk/mailman/listinfo/secnet-announce
61 http://www.chiark.greenend.org.uk/mailman/listinfo/secnet-discuss
63 The -announce list receives one message per secnet release. The
64 -discuss list is for general discussion, including help with
65 configuration, bug reports, feature requests, etc.
67 Bug reports should be sent to <steve@greenend.org.uk>; they will be
68 forwarded to the -discuss list by me.
74 * secnet configuration file format
76 By default secnet on linux reads /etc/secnet/secnet.conf. The default
77 may be different on other platforms.
79 This file defines a dictionary (a mapping from keys to values) full of
80 configuration information for secnet. Two keys must be defined in
81 this file for secnet to start. One is "system", a dictionary
82 containing systemwide control parameters. The other is "sites", a
83 list of all the sites that you intend to communicate with.
85 The configuration file has a very simple syntax; keys are defined as
94 Keys must match the following regular expression:
95 [[:alpha:]_][[:alnum:]\-_]*
97 i.e. the first character must be an alpha or an underscore, and the
98 remaining characters may be alphanumeric, '-' or '_'.
100 Keys can be defined to be a comma-separated list of any of the
106 a dictionary of definitions, enclosed in { }
107 a "closure", followed by arguments
108 a path to a key that already exists, to reference that definition
110 Note that dictionaries can be nested: a key in one dictionary can
111 refer to another dictionary. When secnet looks for a key in a
112 particular directory and can't find it, it looks in the dictionary's
113 lexical 'parents' in turn until it finds it (or fails to find it at
114 all and stops with an error).
116 Definitions can refer to previous definitions by naming them with a
117 path. Paths are key1/key2/key3... (starting from wherever we find
118 key1, i.e. in the current dictionary or any of its parents), or
119 alternatively /key1/key2/key3... (to start from the root).
120 Definitions cannot refer to future definitions.
129 The following paths are valid:
141 Note that f/g/e is NOT 4.
143 Elements that are lists are inserted into lists in definitions, not
144 referenced by them (i.e. you can't have lists of lists).
146 Some closures may be followed by an argument list in ( ), and may
147 return any number of whatever type they like (including other
148 closures). Some types of closure (typically those returned from
149 invokations of other closures) cannot be invoked.
151 closure { definitions } is short for closure({definitions}).
153 The main body of secnet, and all the additional modules, predefine
154 some keys in the root dictionary. The main ones are:
156 yes, true, True, TRUE, on: the boolean value True
157 no, false, False, FALSE, off: the boolean value False
158 makelist: turns a dictionary (arg1) into a list of definitions
160 readfile: reads a file (arg1) and returns it as a string
161 map: applies the closure specified as arg1 to each of the
162 remaining elements in the list in turn. Returns a list
163 made up of the outputs of the closure.
165 Keys defined by modules are described below, in the module
168 Other configuration files can be included inline by writing "include
169 filename" at the start of a line.
171 After the configuration file is read, secnet looks for particular keys
172 in configuration space to tell it what to do:
174 system: a dictionary which can contain the following keys:
175 log (log closure): a destination for system messages
176 userid (string): the userid for secnet to run as once it drops privileges
177 pidfile (string): where to store its PID
179 sites: a list of closures of type 'site', which define other tunnel
180 endpoints that secnet will attempt to communicate with
182 * secnet command line options
184 Usage: secnet [OPTION]...
186 -f, --silent, --quiet suppress error messages
187 -w, --nowarnings suppress warnings
188 -v, --verbose output extra diagnostics
189 -c, --config=filename specify a configuration file
190 -j, --just-check-config stop after reading configfile
191 -n, --nodetach do not run in background
192 -d, --debug=item,... set debug options
193 --help display this help and exit
194 --version output version information and exit
196 * secnet builtin modules
201 adns (closure => resolver closure)
204 config (string): optional, a resolv.conf for ADNS to use
209 randomsrc (closure => randomsrc closure)
211 randomsrc: string[,bool]
212 arg1: filename of random source
213 arg2: if True then source is blocking
218 udp (closure => comm closure)
221 address (string list): IPv6 or IPv4 addresses to listen and send on;
222 default is all local addresses
223 port (integer): UDP port to listen and send on; optional if you
224 don't need to have a stable address for your peers to talk to
225 (in which case your site ought probably to have `local-mobile true').
226 buffer (buffer closure): buffer for incoming packets
227 authbind (string): optional, path to authbind-helper program
232 polypath (closure => comm closure)
234 polypath: dict argument
235 port (integer): UDP port to listen and send on
236 buffer (buffer closure): buffer for incoming packets
237 authbind (string): optional, path to authbind-helper program
238 max-interfaces (number): optional, max number of different interfaces to
239 use (also, maximum steady-state amount of packet multiplication);
240 interfaces marked with `@' do not count.
241 interfaces (string list): which interfaces to process; each entry is
242 optionally `!' or `+' or `@' followed by a glob pattern (which is
243 applied to a prospective interface using fnmatch with no flags).
244 `+' or nothing means to process normally. `!' means to ignore;
245 `@' means to use only in conjunction with dedicated-interface-addr.
246 If no list is specified, or the list ends with a `!' entry, a
247 default list is used/appended:
248 "!tun*","!tap*","!sl*","!userv*","!lo","@hippo*","*".
249 Patterns which do not start with `*' or an alphanumeric need to be
250 preceded by `!' or `+' or `@'.
251 monitor-command (string list): Program to use to monitor appearance
252 and disappearance of addresses on local network interfaces. Should
253 produce lines of the form `+|-<ifname> 4|6 <addr>' where <addr> is
254 an address literal. Each - line should relate to a previously
255 printed + line. On startup, should produce a + line for each
256 currently existing address. secnet does filtering so there is no
257 need to strip out tun interfaces, multicast addresses, and so on.
258 The command is run as the user secnet is started as (not the one
259 which secnet may drop privilege to due to the configured `userid').
260 The default depends on the operating system.
261 permit-loopback (boolean): Normally, loopback IPv6 and IPv4
262 addresses on local interfaces are disregarded, because such
263 interfaces are not interesting for communicating with distant
264 hosts. Setting this option will ignore that check, which can be
265 useful for testing. Setting this option also removes "!lo*" from
266 the default interface pattern list.
268 When using this comm, packets are sent out of every active interface
269 on the host (where possible). It is important that interfaces created
270 by secnet itself are not included! secnet's default filter list tries
273 This comm only makes sense for sites which are mobile. That is, the
274 site closures used with this comm should all have the `local-mobile'
275 parameter set to `true'. When the local site site is not marked
276 mobile the address selection machinery might fixate on an unsuitable
279 polypath takes site-specific informtion as passed to the `comm-info'
280 site closure parameter. The entries understood in the dictionary
282 dedicated-interface-addr (string): IPv4 or IPv6 address
283 literal. Interfaces specified with `@' in `interfaces' will be
284 used for the corresponding site iff the interface local address
287 For an interface to work with polypath, it must either have a suitable
288 default route, or be a point-to-point interface. In the general case
289 this might mean that the host would have to have multiple default
290 routes. However in practice the most useful configuration is two
291 interfaces being (1) wifi (2) mobile internet.
293 I have had success on Linux by using network-manager for wifi and
294 invoking ppp directly for mobile internet. ppp sets up a
295 point-to-point link, and does not add a default route if there already
296 is one. network-manager always sets up a default route. The result
297 is that the wifi always has a default route (so is useable); ppp
298 (being a point-to-point link) does not need one.
300 The use of polypath requires that secnet be started with root
301 privilege, to make the setsockopt(,,SO_BINDTODEVICE,) calls. If the
302 configuration specifies that secnet should drop privilege (see
303 `userid' above), secnet will keep a special process around for this
304 purpose; that process will handle local network interface changes but
305 does not deal with any packets, key exchange, etc.
307 polypath support is only available when secnet is built against an
308 IPv6-capable version of adns (because it wants features in the newer
314 logfile (closure => log closure)
315 syslog (closure => log closure)
317 logfile: dict argument
318 filename (string): where to log to
319 class (string list): what type of messages to log
320 { "debug-config", M_DEBUG_CONFIG },
321 { "debug-phase", M_DEBUG_PHASE },
322 { "debug", M_DEBUG },
323 { "all-debug", M_DEBUG|M_DEBUG_PHASE|M_DEBUG_CONFIG },
325 { "notice", M_NOTICE },
326 { "warning", M_WARNING },
327 { "error", M_ERROR },
328 { "security", M_SECURITY },
329 { "fatal", M_FATAL },
330 { "default", M_WARNING|M_ERROR|M_SECURITY|M_FATAL },
331 { "verbose", M_INFO|M_NOTICE|M_WARNING|M_ERROR|M_SECURITY|M_FATAL },
334 logfile will close and reopen its file upon receipt of SIGHUP.
336 syslog: dict argument
337 ident (string): include this string in every log message
338 facility (string): facility to log as
339 { "authpriv", LOG_AUTHPRIV },
340 { "cron", LOG_CRON },
341 { "daemon", LOG_DAEMON },
342 { "kern", LOG_KERN },
343 { "local0", LOG_LOCAL0 },
344 { "local1", LOG_LOCAL1 },
345 { "local2", LOG_LOCAL2 },
346 { "local3", LOG_LOCAL3 },
347 { "local4", LOG_LOCAL4 },
348 { "local5", LOG_LOCAL5 },
349 { "local6", LOG_LOCAL6 },
350 { "local7", LOG_LOCAL7 },
352 { "mail", LOG_MAIL },
353 { "news", LOG_NEWS },
354 { "syslog", LOG_SYSLOG },
355 { "user", LOG_USER },
361 sysbuffer (closure => buffer closure)
363 sysbuffer: integer[,dict]
366 lockdown (boolean): if True, mlock() the buffer
371 site (closure => site closure)
374 local-name (string): this site's name for itself
375 name (string): the name of the site's peer
376 link (netlink closure)
377 comm (one or more comm closures): if there is more than one, the
378 first one will be used for any key setups initiated by us using the
379 configured address. Others are only used if our peer talks to
381 resolver (resolver closure)
382 random (randomsrc closure)
383 local-key (rsaprivkey closure)
384 address (string list): optional, DNS name(s) used to find our peer;
385 address literals are supported too if enclosed in `[' `]'.
386 port (integer): mandatory if 'address' is specified: the port used
388 key (rsapubkey closure): our peer's public key
389 transform (transform closure): how to mangle packets sent between sites
392 key-lifetime (integer): max lifetime of a session key, in ms
393 [one hour; mobile: 2 days]
394 setup-retries (integer): max number of times to transmit a key negotiation
395 packet [5; mobile: 30]
396 setup-timeout (integer): time between retransmissions of key negotiation
397 packets, in ms [2000; mobile: 1000]
398 wait-time (integer): after failed key setup, wait this long (in ms) before
399 allowing another attempt [20000; mobile: 10000]
400 renegotiate-time (integer): if we see traffic on the link after this time
401 then renegotiate another session key immediately (in ms)
402 [half key-lifetime, or key-lifetime minus 5 mins (mobile: 12 hours),
403 whichever is longer].
404 keepalive (bool): if True then attempt always to keep a valid session key.
406 log-events (string list): types of events to log for this site
407 unexpected: unexpected key setup packets (may be late retransmissions)
408 setup-init: start of attempt to setup a session key
409 setup-timeout: failure of attempt to setup a session key, through timeout
410 activate-key: activation of a new session key
411 timeout-key: deletion of current session key through age
412 security: anything potentially suspicious
413 state-change: steps in the key setup protocol
414 packet-drop: whenever we throw away an outgoing packet
415 dump-packets: every key setup packet we see
416 errors: failure of name resolution, internal errors
417 peer-addrs: changes to sets of peer addresses (interesting for mobile peers)
418 all: everything (too much!)
419 mobile (bool): if True then peer is "mobile" ie we assume it may
420 change its apparent IP address and port number without either it
421 or us being aware of the change; so, we remember the last several
422 port/addr pairs we've seen and send packets to all of them
423 (subject to a timeout). We maintain one set of addresses for key
424 setup exchanges, and another for data traffic. Two communicating
425 peers must not each regard the other as mobile, or all the traffic
426 in each direction will be triplicated (strictly, transmitted
427 mobile-peers-max times) and anyway two peers whose public contact
428 address may suddenly change couldn't communicate reliably because
429 their contact addresses might both change at once. [false]
430 mobile-peers-max (integer): Maximum number of peer port/addr pairs we
431 remember and send to. Must be at least 1 and no more than 5.
432 [4 if any address is configured, otherwise 3]
433 static-peers-max (integer): Maximum number of peer port/addr pairs
434 we can try for a static site. Must be at least 1 and no more
435 than 5. [4 or 3, as above]
436 mobile-peer-expiry (integer): For "mobile" peers only, the length
437 of time (in seconds) for which we will keep sending to multiple
438 address/ports from which we have not seen incoming traffic. [120]
439 local-mobile (bool): if True then other peers have been told we are
440 "mobile". This should be True iff the peers' site configurations
441 for us have "mobile True" (and if we find a site configuration for
442 ourselves in the config, we insist on this). The effect is to
443 check that there are no links both ends of which are allegedly
444 mobile (which is not supported, so those links are ignored) and
445 to change some of the tuning parameter defaults. [false]
446 mtu-target (integer): Desired value of the inter-site MTU for this
447 peering. This value will be advertised to the peer (which ought
448 to affect incoming packets), and if the peer advertises an MTU its
449 value will be combined with this setting to compute the inter-site
450 MTU. (secnet will still accept packets which exceed the
451 (negotiated or assumed) inter-site MTU.) Setting a lower
452 inter-site MTU can be used to try to restrict the sizes of the
453 packets sent over the underlying public network (e.g. to work
454 around network braindamage). It is not normally useful to set a
455 larger value for mtu-target than the VPN's general MTU (which
456 should be reflected in the local private interface MTU, ie the mtu
457 parameter to netlink). If this parameter is not set, or is set
458 to 0, the default is to use the local private link mtu.
459 comm-info (dict): Information for the comm, used when this site
460 wants to transmit. If the comm does not support this, it is
463 Links involving mobile peers have some different tuning parameter
464 default values, which are generally more aggressive about retrying key
465 setup but more relaxed about using old keys. These are noted with
466 "mobile:", above, and apply whether the mobile peer is local or
472 eax-serpent (closure => transform closure)
477 serpent256-cbc (closure => transform closure)
482 null-netlink (closure => closure or netlink closure)
484 null-netlink: dict argument
485 name (string): name for netlink device, used in log messages
486 networks (string list): networks on the host side of the netlink device
487 remote-networks (string list): networks that may be claimed
488 by the remote site using this netlink device
489 local-address (string): IP address of host's tunnel interface
490 secnet-address (string): IP address of this netlink device
491 ptp-address (string): IP address of the other end of a point-to-point link
492 mtu (integer): MTU of host's tunnel interface
494 Only one of secnet-address or ptp-address may be specified. If
495 point-to-point mode is in use then the "routes" option must also be
496 specified, and netlink returns a netlink closure that should be used
497 directly with the "link" option to the site closure. If
498 point-to-point mode is not in use then netlink returns a closure that
499 may be invoked using a dict argument with the following keys to yield
501 routes (string list): networks reachable down the tunnel attached to
502 this instance of netlink
503 options (string list):
504 allow-route: allow packets coming from this tunnel to be routed to
505 other tunnels as well as the host (used for mobile devices like laptops)
506 soft: remove these routes from the host's routing table when
507 the tunnel link quality is zero
508 mtu (integer): MTU of host's tunnel interface
510 Netlink will dump its current routing table to the system/log on
516 userv-ipif (closure => netlink closure)
518 userv-ipif: dict argument
519 userv-path (string): optional, where to find userv ["userv"]
520 service-user (string): optional, username for userv-ipif service ["root"]
521 service-name (string): optional, name of userv-ipif service ["ipif"]
522 buffer (buffer closure): buffer for assembly of host->secnet packets
523 plus generic netlink options, as for 'null-netlink'
528 tun (closure => netlink closure) [only on linux-2.4]
529 tun-old (closure => netlink closure)
532 flavour (string): optional, type of TUN interface to use
533 ("guess","linux","bsd","streams")
534 device (string): optional, path of TUN/TAP device file ["/dev/net/tun"]
535 interface (string): optional, name of tunnel network interface
536 ifconfig-path (string): optional, path to ifconfig command
537 route-path (string): optional, path to route command
538 ifconfig-type (string): optional, how to perform ifconfig
539 route-type (string): optional, how to add and remove routes
540 types are: "guess", "ioctl", "bsd", "linux", "solaris-2.5"
541 buffer (buffer closure): buffer for host->secnet packets
542 plus generic netlink options, as for 'null-netlink'
544 I recommend you don't specify the 'interface' option unless you're
545 doing something that requires the interface name to be constant.
550 rsa-private (closure => rsaprivkey closure)
551 rsa-public (closure => rsapubkey closure)
553 rsa-private: string[,bool]
554 arg1: filename of SSH private key file (version 1, no password)
555 arg2: whether to check that the key is usable [default True]
557 rsa-public: string,string
558 arg1: encryption key (decimal)
559 arg2: modulus (decimal)
564 diffie-hellman (closure => dh closure)
566 diffie-hellman: string,string[,bool]
568 arg2: generator (hex)
569 arg3: whether to check that the modulus is prime [default True]
584 makelist (dictionary => list of definitions)
585 readfile (string => string)
586 map (closure,list => list)
589 returns a list consisting of the definitions in the dictionary. The keys
593 reads the named file and returns its contents as a string
596 applies the closure specified as arg1 to each of the elements in the list.
597 Returns a list made up of the outputs of the closure.