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