1 * Thu Nov 17 2005 Brendan O'Dea <bod@optus.net> 2.1.12
2 - Set MTU on tunnel interface so the kernel will re-fragment large
4 - Fix TCP checksum recalc.
5 - NAK silly MRU values for peer.
7 * Mon Nov 14 2005 Brendan O'Dea <bod@optus.net> 2.1.11
8 - Fix fragment handling in ip_filter.
9 - Exclude counter when comparing filter rules.
11 * Sat Nov 5 2005 Brendan O'Dea <bod@optus.net> 2.1.10
12 - Add scripts/l2tpns-capture.
13 - Fix LCP Echo frequency.
14 - Add Framed-Route entries to RADIUS records.
15 - Reset restart counters correctly.
16 - Reset timers on sending ConfigReq.
17 - Only send one RADIUS Start record, even if IPCP is restarted.
19 * Tue Oct 11 2005 Brendan O'Dea <bod@optus.net> 2.1.9
20 - Fix Calling-Station-Id in RADIUS accounting records (Slobodan Tomic).
21 - Fix RADIUS authentication on DAE responses.
22 - Don't send tunnel HELLO when there are pending control messages.
23 - Move plugin_radius_reset from *ctl to auto* plugins.
24 - Add Cisco-AVPairs to RADIUS accounting records via plugin_radius_account.
26 * Mon Sep 19 2005 Brendan O'Dea <bod@optus.net> 2.1.8
27 - Move code from signal handlers into mainloop, avoiding a race
28 condition when forking CLI.
30 * Fri Sep 16 2005 Brendan O'Dea <bod@optus.net> 2.1.7
31 - This time, for sure: really fix Protocol-Reject.
33 * Fri Sep 16 2005 Brendan O'Dea <bod@optus.net> 2.1.6
34 - Any traffic on a tunnel resets lastrec, not just control messages.
35 - Use a unique identifier for LCP.
36 - Fix Code-Reject/Protocol-Reject.
37 - Add l2tp_mtu configuration option, used to define MRU, MSS.
38 - Adjust TCP MSS options in SYN and SYN,ACK packets to avoid
39 fragmentation of tcp packets.
41 * Sat Sep 3 2005 Brendan O'Dea <bod@optus.net> 2.1.5
42 - Avoid Code-Reject loop.
43 - Increase size of PPP buffers to MAXETHER.
44 - Bug fixes for CLI ringbuffer and tunnel HELLO from Yuri.
45 - Restart rather than halt BGP on receipt of CEASE (Dominique Rousseau).
46 - Add cluster_mcast_ttl option to allow a cluster to span multiple
47 subnets (suggested by Tim Devries).
49 * Mon Aug 29 2005 Brendan O'Dea <bod@optus.net> 2.1.4
50 - Drop level of "Unexpected CHAP message" log.
51 - Fix parsing of ProtocolRej (allow 1 or two byte protocols).
52 - Handle rejection of MRU negotiation by peer.
53 - Use local hostname for tunnel in SCCRP (Alex Kiernan).
55 * Wed Aug 17 2005 Brendan O'Dea <bod@optus.net> 2.1.3
56 - Fail IPCP negotiation only on ConfigRej of IP-Address.
58 * Wed Aug 10 2005 Brendan O'Dea <bod@optus.net> 2.1.2
59 - Clear cluster_master on election so that slaves will accept a new master.
60 - Provide more comments/defaults in etc/startup-config.default.
61 - Add DAE support (PoD/CoA) from Vladislav Bjelic.
62 - Clean up new warnings from gcc 4.0.
63 - Replace flags used for LCP/IPCP with state machine.
64 - Include Acct-Session-Time in interim records.
65 - Fix DAE vector, generateload (Alex Kiernan).
66 - Replace RSA MD5 with public domain version.
68 * Tue Jun 14 2005 Brendan O'Dea <bod@optusnet.com.au> 2.1.1
69 - Add missing newline to backtrace macro.
70 - Don't send CDN for each session when shutting down tunnels (this is
72 - Move tunnel shutdown from SIGQUIT signal handler to be run once from
73 still_busy(). Reject new tunnels/sessions while in the process of
75 - Clarify usage of shutdown signals in documentation.
76 - Always initialise PRNG.
77 - Sanity check length of random_vector.
78 - Fix segv in unhide_value.
79 - Ping new master when we get C_MASTER and delay next election to allow
80 the unicast limp-along code to kick in if required.
82 * Sun Jun 5 2005 Brendan O'Dea <bod@optusnet.com.au> 2.1.0
83 - Add IPv6 support from Jonathan McDowell.
84 - Add CHAP support from Jordan Hrycaj.
85 - Add interim accounting support from Vladislav Bjelic.
86 - Negotiate MRU, default 1458 to avoid fragmentation.
87 - Sanity check that cluster_send_session is not called from a child
89 - Use bounds-checking lookup functions for string constants.
90 - Add enum for RADIUS codes.
91 - Make "call_" prefix implict in CSTAT() macro.
92 - Fix some format string problems.
93 - Remove "save_state" option. Not maintained anymore; use clustering
94 to retain state across restarts.
95 - Simplify AVP unhiding code.
96 - Add optional "username" parameter to ungarden control, allowing the
97 username to be reset before going online.
98 - Add result/error codes and message to StopCCN when shutting down tunnels.
99 - Add result/error codes to CDN when shutting down sessions. Sends 2/7
100 (general error, try another LNS) when out of IP addresses, and 3
101 (adminstrative) for everything else (suggestion from Chris Gates).
102 - Use cli_error() for error messages and help.
103 - Don't use LOG() macro in initdata() until the config struct has been
104 allocated (uses config->debug).
105 - Initialise log_stream to stderr to catch errors before the config file
107 - Make "show running-config" a privileged command (contains clear text
109 - Add sessionctl plugin to provide drop/kill via nsctl.
110 - New config option: allow_duplicate_users which determines whether
111 or not to kill older sessions with the same username.
112 - Fix byte counters in accounting records.
113 - Add Acct-Output-Gigawords, Acct-Input-Gigawords attributes to RADIUS
115 - Fix icmp host unreachable to use router address.
116 - Include endpoint address in accounting dump files.
117 - Convert mainloop to use epoll rather than select.
118 - Add note about fragmentation in Docs/manual.html, and a sample
119 iptables rule for MSS clamping.
121 + Show session open time in "show session"/"show user" detailed output.
122 + Have slaves with BGP configured drop BGP on receipt of a shutdown
123 signal, but hang about for an additional 5s to process any remaining
125 + Run regular_cleanups after processing the results of the select,
126 looking at a sufficient slice of each table to ensure that all
127 entries are examined at least once per second.
129 + Cluster changes from Michael, intended to prevent a stray master
130 from trashing a cluster:
131 = Ignore heartbeats from peers claiming to be the master before the
132 timeout on the old master has expired.
133 = A master receiving a stray heartbeat sends a unicast HB back, which
134 should cause the rogue to die due to the tie-breaker code.
135 = Keep probing the master for late heartbeats.
136 = Drop BGP as soon as we become master with the minumum required peers.
137 = Any PING seen from a master forces an election (rather than just
138 where basetime is zero).
139 = A slave which receives a LASTSEEN message (presumably a restarted
140 master) sends back new message type, C_MASTER which indicates the
141 address of the current master.
142 + New config option: cluster_master_min_adv which determines the minimum
143 number of up to date slaves required before the master will drop
146 + Add handling of "throttle=N" RADIUS attributes.
147 + Fix RADIUS indexing (should have 16K entries with 64 sockets).
149 + Fix leak in session freelist when initial RADIUS session allocation
152 + Add a Cisco-Avpair with intercept details to RADIUS Start/Stop
155 + Only send RADIUS stop record in sessionshutdown when there's an ip address.
156 + Reset .die on master takeover (so that dying sessions don't have to
157 hang around until the new master has the same uptime as the old one).
158 + Update .last_packet in cluster_handle_bytes only when there have
159 been bytes received from the modem (dead sessions were having the
160 idle timeout reset by stray packets).
162 + Ensure that sessionkill is not called on an unopened session (borks
164 + Bump MAXSESSION to 60K.
165 + Fix off-by-one errors in session/tunnel initialisation and
166 sessiont <-> sessionidt functions.
167 + Use session[s].opened consistently when checking for in-use sessions
168 (rather than session[s].tunnel).
169 + Use <= cluster_highest_sessionid rather than < MAXSESSION in a
171 + Don't kill a whole tunnel if we're out of sessions.
172 + Change session[s].ip to 0 if set from RADIUS to 255.255.255.254;
173 avoids the possibility that it will be interpreted as a valid IP
175 + Avoid a possible buffer overflow in processpap.
176 + Kill session if authentication was rejected.
178 + More DoS prevention: add packet_limit option to apply a hard limit
179 to downstream packets per session.
180 + Fix "clear counters".
181 + Log "Accepted connection to CLI" at 4 when connection is from localhost
182 to reduce noise in logs.
183 + Show time since last counter reset in "show counters".
185 + Throttle outgoing LASTSEEN packets to at most one per second for a
188 * Fri Dec 17 2004 Brendan O'Dea <bod@optusnet.com.au> 2.0.13
189 - Better cluster master collision resolution: keep a counter of state
190 changes, propagated in the heartbeats; the master with the highest #
191 of changes (that has kept in contact with the LAC through the
193 - Skip newlines in ringbuffer messages to CLI.
194 - Drop "Session N is closing" message level to 4; don't process PPPIP
195 packets in this state.
196 - Use gzip --best for man pages, include pid_file in sample
197 startup-config (from Jonathan's Debian package patches).
198 - Read multiple packets off cluster_sockfd as well as udpfd, tunfd in an
199 attempt to avoid losing the cluster in high load (DoS) conditions.
200 - Add counters for select_called, multi_read_used and multi_read_exceeded.
202 - Retain counters of shutdown sessions to dump once per minute.
203 - Use standard uintN_t types for portability.
205 * Wed Dec 1 2004 Brendan O'Dea <bod@optusnet.com.au> 2.0.12
206 - The "This time, for sure!" release.
207 - Fix throttlectl plugin argument parsing.
209 * Wed Dec 1 2004 Brendan O'Dea <bod@optusnet.com.au> 2.0.11
210 - Don't send a RADIUS start record when ungardening on shutdown.
212 * Wed Dec 1 2004 Brendan O'Dea <bod@optusnet.com.au> 2.0.10
213 - Fix byte ordering of LCP header length (thanks Yuri).
214 - Increase ip_conntrack_max due to dropped packets.
216 * Tue Nov 30 2004 Brendan O'Dea <bod@optusnet.com.au> 2.0.9
217 - Revise CCP, send ConfigReq once only.
218 - Don't copy the old buffer into Config{Nak,Rej} LCP responses (oops);
219 add length checks when appending.
220 - Do copy the identifier from the request and update length.
221 - Have makeppp print a backtrace on overflow.
222 - Check control serial before clearing window, prevents looping tunnel
223 setup in some instances.
224 - Implement named access-lists which may be applied to a session
225 either via Filter-Id RADIUS responses, or using the CLI.
226 - Drop ip address from LOG.
227 - autothrottle: revise parsing; ignore lcp:interface-config avpairs
228 which don't start with serv[ice-policy].
231 * Sat Nov 20 2004 Brendan O'Dea <bod@optusnet.com.au> 2.0.8
232 - Ignore gateway address in Framed-Route (from Jonathan McDowell).
233 - Don't route Framed-IP-Address if contained in a Framed-Route.
234 - Call sessionshutdown() when a tunnel is dropped rather than
235 sessionkill() to ensure that RADIUS stop records are sent.
236 - Cleanup: make a bunch of global functions/variables static.
237 - Remove reference to old -a command line argument.
238 - Add l2tpns(8) and nsctl(8) manpages from Jonathan McDowell.
239 - Add startup-config(5) manpage.
240 - Revise nsctl to allow arbitrary strings/args to be passed to plugins.
241 - Add snoopctl, throttlectl plugins.
242 - Fix deletion from linked list.
243 - Allow LCP re-negotiation after connection completes (thanks Yuri).
245 * Mon Nov 15 2004 Brendan O'Dea <bod@optusnet.com.au> 2.0.7
246 - Fix socket creation in host_unreachable() (thanks to Bjørn Augestad)
247 - Don't assume BGP peer sends back negotiated hold time, pick smallest
249 * Thu Nov 11 2004 Brendan O'Dea <bod@optusnet.com.au> 2.0.6
250 - Make BGP keepalive/hold time configurable
251 - Revise BGP config to use "router bgp AS" syntax (requires libcli >= 1.8.2)
253 * Tue Nov 9 2004 Brendan O'Dea <bod@optusnet.com.au> 2.0.5
254 - Handle routing properly in lone-master case
255 - Fix intercepts: don't double-snoop throttled customers, ensure
256 byte/packet counts are only updated once
257 - Add a callback to allow plugins to fetch values from the running config
259 * Mon Nov 8 2004 Brendan O'Dea <bod@optusnet.com.au> 2.0.4
260 - Added setrxspeed plugin
261 - Added peer_address config option
262 - Rename wrapper macros to LOG()/LOG_HEX(), use p->log() in plugins
263 - Replace some PPP{PAP,CHAP} magic numebrs with constants
264 - Nak asyncmap (unless == 0)
265 - Bundle ConfigRej options
266 - Clean up initlcp handling
268 * Wed Nov 3 2004 Brendan O'Dea <bod@optusnet.com.au> 2.0.3
269 - Added support for hidden AVPs by Robert Clark
270 - l2tpns-chap-response.patch from Robert Clark
271 - Merge l2tpns-config-hostname.patch from Robert Clark
272 - l2tpns-dont-timeshift-unidirectional-traffic.patch from Robert Clark - Dump accounting data if cin OR cout is non-zero
273 - Don't write accounting files if no accounting dir is set - Yuri
274 - Fix checking for mmap success
275 - Renegotiate MRU - Yuri
276 - Take LCP ConfigReq length from the packet length field - Yuri
277 - Hostname set via command line not config
278 - Make number of throttle buckets configurable
279 - Shared_malloc returns NULL on failure
281 - Unfsck 4->8 indenting change
282 - Use 2 seperate u16 values for throttle rate in/out
283 - Defer adding radius fds to the select loop until become_master
285 * Thu Sep 02 2004 David Parrish <david@dparrish.com> 2.0.2
286 - Combined LCP patches from Iain and Yuri. This should allow Windows 2k/XP
287 clients to connect, as well Linksys DSL modems.
288 - Apply patch to fix -v option from Juergen Kammer.
289 - Makefile fix from Juergen Kammer to not overwrite existing config files on
291 - Configurable radius port patch from Juergen Kammer.
292 - Send my_address if no bind_address when doing IPCP
293 - Write pid file if filename is set
294 - Add startup script and monitor script from Yuri
295 - Some logging correctness fixes from Iain Wade
296 - Add support for LCP Ident and CallBack (rejection only) from Yuri
297 - Initiate LCP if not attempted by the client, or in renegotiation - Yuri
298 - Indentation and style cleanups
299 - Per-user upload and download throttle rates - Yuri
300 - Make autothrottle.so understand cisco lcp:interface-config - Yuri
301 - Show filter stats in show session - Yuri
302 - Cleanup from Michael to change sid to unique_id
303 - Add plugin to remove domain name from auth requests
304 - Add .spec file for RPM generation
306 * Tue Jul 13 2004 Brendan O'Dea <bod@optusnet.com.au> 2.0.1
307 - Update INSTALL, Docs/manual.html documentation.
308 - Add INTERNALS documentation.
309 - Add lock_pages option.
310 - TerminateAck fix from Yuri
311 - Adject cli_loop args for libcli 1.8.0
312 - Allow for backward compatabity in C_PING packets
313 - Don't send RADIUS stop messages from sessionshutdown when called from
316 - Fix for LASTSEEN breakage: don't do anything in the CLI other than
317 flag changes to be made by the parent.
318 - Split out master parts from cluster_check_master() into cluster_check_slaves().
319 - Set hostname in CLI prompt.
320 - Make cluster_hb_interval work; include interval/timeout in heartbeats
321 so that a change on the master is propagated immediately to the slaves.
322 - Use fast heartbeats when there are slaves not up to date.
323 - Ensure basetime of shut down master is set to zero (prevent delayed election).
324 - Fix radius session leak on IPCP timeout.
325 - Fix some off-by-one errors in tunnel/session loops.
326 - Add "limp along" fix for when a slave drops temporarily from the mcast group.
327 - Rename l2tpns.cfg as startup-config to match CONFIGFILE.
328 - Update cli callbacks to work with libcli 1.6.
329 This supports privileged and unprivileged commands, as well as a configuration
331 - Add help for all cli commands.
332 - Add "show version" command.
333 - Fix uptime counter display.
334 - Fix nasty bug where cluster basetime can be set to 0 when sending initial
336 - Don't rmmod ip_conntrack, as this can take a lot of time.
337 - Re-order logging in routeset such that the action is given before any error.
338 - Use the correct gateway address when deleting routes.
339 - Remove any routes when address changes.
340 - Require authentication if telnet from remote ip.
341 - Require enable password always.
342 - Return error if show pool done on slave.
343 - We MUST immediately exit if we're the wrong master!
345 * Wed Jun 23 2004 David Parrish <david@dparrish.com> 2.0.0
347 - Completely replace active/standby clustering with a new peer-to-peer
348 clustering method which allows much greater throughput and is a lot more fault
350 - Add internal tbf implementation for throttling without relying on tc and
352 - Add support for iBGP and eBGP to advertise routes
353 - Add cli commands "show cluster", "show bgp", "show ipcache", "show throttle",
354 "show tbf", "suspend bgp", "restart bgp", "show user"
355 - Interception destination must be set per-user
356 - If SMP machine, allow use of SCHED_FIFO, which should improve performance
357 - Added config option to send GARP at startup
358 - Added plugin_become_master and plugin_new_session_master plugin hooks
359 - Remove useless sessionsendarp(). This isn't needed now that we are using TUN
361 - ICMP rate limiting so not every unreachable packet is replied with an ICMP
363 - mangle table is not required on anything but the cluster master, so slaves
364 will drop the mangle table and attempt to unload the ip_conntrack module
365 - Statically assigned IP addresses (by Radius) work now
366 - Add -d command-line flag to detach and become a daemon
367 - Configuration file is now "/etc/l2tpns/startup-config"
368 - Reduced MIN_IP_SIZE to 0x19 to stop a pile of Short IP warnings
369 - Resend initial IPCP request until it's acknowleged by the client
370 - Better radius session cleanup logic
371 - Many miscellaenous bugfixes and performance enhancements
372 - Thanks to Michael O'Reilly and Brendan O'Dea for most of these new features
374 * Mon May 24 2004 David Parrish <david@dparrish.com> 1.2.0
375 - Fix SEGFAULT in garden module
376 - Use multiple radius sockets to allow more concurrent authentication requests
377 - Add username parameter to "show users" command
378 - Fix counting tunnel rx errors as tunnel tx errors
379 - Add "show throttle" command
380 - Add gcc __attribute__ to logging functions
381 - Fix warnings shown by __attribute__
382 - Make sure regular cleanup happens regularly under high load
383 - Add variable cleanup_interval for changing cleanup interval
384 - Add support for reading more than one packet per fd in each processing loop
385 - This is configurable with the multi_read_count variable
386 - Remove segv handler so core dumps can happen
387 - Use nonblocking sockets
388 - Increase tun queue length
389 - Fix minimum length of IP packets
390 - Remove per-packet plugin hooks (they are slow)
391 - Don't drop session if no free RADIUS
392 - Don't expire more than 1000 sessions per cleanup interval
393 - Remove -a and -c command-line options. They don't work anyway
394 - Don't require file: in log_filename
395 - Bump version to 1.2.0
396 - Check return code when throttling users
398 * Mon Apr 5 2004 David Parrish <david@dparrish.com> 1.1.1
399 - Don't mention configure anymore, it's not used
400 - Added the autosnoop and autothrottle modules
401 - Don't default to using a htb for the class root
403 * Fri Mar 5 2004 David Parrish <david@dparrish.com> 1.1.0
404 - Change all strcpy() calls to strncpy() to avoid buffer overflow potential
405 - Add ICMP host unreachable support
406 - Logging to syslog if log_file = "syslog:facility"
407 - Now requires libcli 1.5
408 - All configuration moves to a config structure
409 - Ability to modify and write config on the fly through command-line interface
410 - Config file support is removed, and now handled by the cli
411 - Show hostname in cli prompt
412 - Keep current state type for tunnels
413 - Add uptime command do CLI, which also shows real-time bandwidth utilisation
414 - Add goodbye command to cluster master, which forces droppping a slave
415 - Cache IP address allocation, so that reconnecting users get the same address
416 - Fix tunnel resend timeouts, so that dead tunnels will be cleaned up
417 - Allocate tunnels and radius without using a linked list which had issues
418 - Fix some off-by-one errors in tunnel and session and radius arrays
419 - Save and reload ip address pool when dieing
420 - Check version and size of reloaded data when restarting
421 - Remove plugin_config support
422 - Remove old support for TBF which didn't work anyway. HTB is required to do throttling now.
423 - Add COPYING and Changes files