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