1 %define initdir %{_sysconfdir}/init.d
5 Packager: Samba Team <samba@samba.org>
9 License: GNU GPL version 3
10 Group: System Environment/Daemons
11 URL: http://ctdb.samba.org/
13 Source: ctdb-%{version}.tar.gz
15 Prereq: /bin/mktemp /usr/bin/killall
16 Prereq: fileutils sed /etc/init.d
17 Prereq: /usr/bin/tdbdump /usr/bin/tdbtool
19 Provides: ctdb = %{version}
22 BuildRoot: %{_tmppath}/%{name}-%{version}-root
25 ctdb is the clustered database used by samba
28 #######################################################################
34 # setup the init script and sysconfig file
35 %setup -T -D -n ctdb-%{version} -q
40 if ccache -h >/dev/null 2>&1 ; then
48 ## always run autogen.sh
51 CFLAGS="$RPM_OPT_FLAGS $EXTRA -O0 -D_GNU_SOURCE -DCTDB_VERS=\"%{version}-%{release}\"" ./configure \
53 --sysconfdir=%{_sysconfdir} \
55 --localstatedir="/var"
57 make docdir=%{_docdir} showflags
58 make docdir=%{_docdir}
61 # Clean up in case there is trash left from a previous build
62 rm -rf $RPM_BUILD_ROOT
64 # Create the target build directory hierarchy
65 mkdir -p $RPM_BUILD_ROOT%{_sysconfdir}/sysconfig
66 mkdir -p $RPM_BUILD_ROOT%{_sysconfdir}/init.d
68 make DESTDIR=$RPM_BUILD_ROOT docdir=%{_docdir} install
70 install -m644 config/ctdb.sysconfig $RPM_BUILD_ROOT%{_sysconfdir}/sysconfig/ctdb
71 install -m755 config/ctdb.init $RPM_BUILD_ROOT%{initdir}/ctdb
73 mkdir -p $RPM_BUILD_ROOT%{_docdir}/ctdb/tests/bin
74 install -m755 tests/bin/ctdb_transaction $RPM_BUILD_ROOT%{_docdir}/ctdb/tests/bin
76 # Remove "*.old" files
77 find $RPM_BUILD_ROOT -name "*.old" -exec rm -f {} \;
80 rm -rf $RPM_BUILD_ROOT
83 #######################################################################
85 #######################################################################
90 %config(noreplace) %{_sysconfdir}/sysconfig/ctdb
91 %config(noreplace) %{_sysconfdir}/ctdb/notify.sh
92 %config(noreplace) %{_sysconfdir}/ctdb/ctdb-crash-cleanup.sh
93 %config(noreplace) %{_sysconfdir}/ctdb/functions
94 %attr(755,root,root) %{initdir}/ctdb
96 %{_docdir}/ctdb/README.eventscripts
97 %{_docdir}/ctdb/recovery-process.txt
98 %{_sysconfdir}/ctdb/events.d/00.ctdb
99 %{_sysconfdir}/ctdb/events.d/01.reclock
100 %{_sysconfdir}/ctdb/events.d/10.interface
101 %{_sysconfdir}/ctdb/events.d/13.per_ip_routing
102 %{_sysconfdir}/ctdb/events.d/11.natgw
103 %{_sysconfdir}/ctdb/events.d/11.routing
104 %{_sysconfdir}/ctdb/events.d/20.multipathd
105 %{_sysconfdir}/ctdb/events.d/31.clamd
106 %{_sysconfdir}/ctdb/events.d/40.vsftpd
107 %{_sysconfdir}/ctdb/events.d/41.httpd
108 %{_sysconfdir}/ctdb/events.d/50.samba
109 %{_sysconfdir}/ctdb/events.d/60.nfs
110 %{_sysconfdir}/ctdb/events.d/60.ganesha
111 %{_sysconfdir}/ctdb/events.d/62.cnfs
112 %{_sysconfdir}/ctdb/events.d/70.iscsi
113 %{_sysconfdir}/ctdb/events.d/91.lvs
114 %{_sysconfdir}/ctdb/statd-callout
115 %{_sysconfdir}/ctdb/interface_modify.sh
120 %{_bindir}/ctdb_diagnostics
122 %{_mandir}/man1/ctdb.1.gz
123 %{_mandir}/man1/ctdbd.1.gz
124 %{_mandir}/man1/onnode.1.gz
125 %{_libdir}/pkgconfig/ctdb.pc
127 %{_docdir}/ctdb/tests/bin/ctdb_transaction
130 Summary: CTDB developmnet libraries
134 development libraries for ctdb
137 %defattr(-,root,root)
138 %{_includedir}/ctdb.h
139 %{_includedir}/ctdb_client.h
140 %{_includedir}/ctdb_protocol.h
141 %{_includedir}/ctdb_private.h
142 %{_includedir}/ctdb_typesafe_cb.h
146 * Mon Jun 27 2011 : Version 1.2.27-202
147 - Remove an annoying log message
149 * Mon May 30 2011 : Version 1.2.27-201
150 - Remove collection of GPFS data from ctdb-diagnostics
152 - Always try to restore all routes regardless of which (VLAN) interface
153 that addresses were added/removed from.
155 - Verify a pointer before dereference
157 - Restart samba if it failed to start
159 * Tue May 17 2011 : Version 1.2.27-200
160 - rename to -200 so we leave some space for non-collission with the PTF1
162 * Tue May 10 2011 : Version 1.2.27-3
163 - Add support for bonding mode 4, workaround thel5 kernel bug
164 - Restart samba if it fails 'service smb start' CQ S1023394
165 - Dont stop checking interfaces when deleted interfce found : CQ S1023420
166 * Mon Apr 11 2011 : Version 1.2.27-2
167 - Update and simplify handling of interfaces.
168 check for and become unhealthy if an interface has been deleted
169 * Mon Mar 21 2011 : Version 1.2.27
170 - Fix a tdb_lock all bug in the ctdb_req_dmaster from non-master code.
172 * Mon Mar 21 2011 : Version 1.2.26
173 - Fix a talloc hierarchy bug for the deferred attach context
175 * Mon Mar 14 2011 : Version 1.2.25
176 - Vacuum update. Vacuuming updated from Michael Adam's vacuuming patch which
177 greatly improves the speed of vacuuming of shortlived records.
178 - Defer database access until we have well gone through initial startup and covergence. This reduces the chance of hitting a split brain when writing to the cluster if we dont use split brain protection reclock file.
179 - Try to restart recovery daemon if/when it fails or hangs.
180 * Fri Feb 25 2011 : Version 1.2.23
181 - Fix for "DB Attach to database %s refused. Can not match clientid.."
182 that was introduced in 1.2.22
183 * Wed Feb 23 2011 : Version 1.2.22
184 - Fix transaction errors that especially happened during startup.
185 * Mon Feb 21 2011 : Version 1.2.21
187 * Tue Feb 8 2011 : Version 1.2.20
188 - Make addip command more robust for "slow" systems.
189 * Mon Jan 31 2011 : Version 1.2.19
190 - Extra logging to show why CTDB dont start if it is in STOPPED state
192 - Collapse all DMASTER migration requests on one single dmaster node for a record into one single lockwait child.
194 * Wed Jan 18 2011 : Version 1.2.18
195 - Fix for a deadlock in the TDB library, causing "Time Jumped" error messages
196 and hung CTDBD process.
197 * Mon Jan 17 2011 : Version 1.2.17
198 - Updates to the event logging from Christian.
199 - Libctdb : traverse support
200 - Updates to the event logging by adding internal ticker.
201 - Addip/Delip/Moveip : make more robust by adding retry if failure
202 - Dont ban nodes for ipreallocation failures, being unhealthy is sufficient
203 - NFS eventscripts : lower the fail/restart limits
204 - Eventscript restart : only restart on monitor event.
205 - Libctdb crash issue from Mike Anderson
206 - chante daemon back to real-time process to handle high load better and reduce
207 time jumped messages.
208 - nfs eventscript : skip checking statd when running on SLES systems
209 * Wed Dec 22 2010 : Version 1.2.16
210 - 50.samba dont run serverid wipe in the background in case it
211 is so slow to start that samba manages to come up before it finishes.
212 - 60.nfs wait 10 intervals before trying to restart lockd.
213 flag the node unhealthy after 15 failures.
215 - 41.httpd httpd can sometimes be slow, wait 5 intervals before we try to
216 restart it and 10 intervals before we flag the node unhealthy.
217 * Mon Dec 13 2010 : Version 1.2.15
218 - Add two new debugging commands "ctdb readkey/writekey"
219 - idtree overflow bugfix
220 - only run "serverid wipe" when we are actually running samba
221 - libctdb, add roper input queueing so we can support calling
222 sync functions from an async callback
224 - addip, always wait across at least one ip reallocation, making the
225 command slower, but making it easier to use in scripts
226 * Mon Dec 6 2010 : Version 1.2.14
227 - detect when ip allocation is inconsistent and some hosts show -1
228 as the node for the ip.
230 - nfs ganesha support in eventscripts
231 - remove lacount/laccessor from the ltdb header
232 * Fri Dec 3 2010 : Version 1.2.13
233 - Fix a bug with the script autostart feature.
234 Dont try to start samba during the init event
235 * Mon Nov 29 2010 : Version 1.2.12
236 - If we are no longer natgw master, tell winbind to recycle the socket.
237 * Mon Nov 22 2010 : Version 1.2.11
238 - Fix bugs in the evenstcripts where we are supposed to automatically
239 start/stop services when CTDB_MANAGES_* changes, and we were not.
240 - Change 60.nfs to allow kNFSd to have been dead for quite a long time
241 before we flag the node as unehalthy and failover.
242 * Mon Nov 15 2010 : Version 1.2.10
243 - Make sure to initialize the statistics start time to current time
244 instead of leaving it to point to start of epoch.
246 - Create a new tunable DisableIPFailover that is used to tell ctdb
247 to not check any ip allocation at all and never do any failover
248 This can be used to stop/restart individual nodes without causing
249 any ip failovers to happen.
250 * Tue Nov 2 2010 : Version 1.2.9
251 - Drop loglevels on several items and remove spam from the messages file
252 - Both nfs and nfslock can fail so restart both if there is a problem
253 * Mon Oct 25 2010 : Version 1.2.8
254 - Allow samba to specify that a new database to attach to/ create
255 should use Jenkins3 hash : CQ1019744
256 * Mon Oct 18 2010 : Version 1.2.7
257 - Dont monitor GPFS filesystems in 62.cnfs
258 - If tdb_open() fails, print errno to make troubleshooting easier
259 - Try restarting RPC.LOCKD if it failed to start
260 - Remove a dbug message
261 - Make sure the statd state directory exists before trying to touch files in it.
262 * Tue Oct 12 2010 : Version 1.2.6
263 - Move config.tdb handling into a function in 00.ctdb
264 - Latency counters min/max/avg for all latency statistics
265 - Update default hash size to be 100001
266 - Check all bond devices, dont exit after the first one
267 - Change to useing the Jenkins hash for LMASTER selection
268 - Sync with TDB from upstream samba
270 - remove some log messages
271 - add rolling statistics
273 * Thu Sep 9 2010 : Version 1.2.5
274 - Suppress some VSFTPD warnings
275 - Make sure all STATD directories exist before we dereference them
277 - Fix for a crash when we write a debug message after a memory allocation
278 fail. Fix the message and call ctdb_fatal() properly
279 - Move the state directory off /etc/ctdb and to /var/ctdb
280 - Natgw changes to allow "slave only" natgw members
281 - Fix "ctdb listnodes" so it works again.
282 * Mon Aug 30 2010 : Version 1.2.4
283 - fix a bug in the ipreallocate eventscript handling
284 - remove a debug logentry
285 * Mon Aug 30 2010 : Version 1.2.3
286 - finish moving CTDB off the udnerlying filesystem.
287 both tickles and statd information
288 - new infrastructure to support the above
289 - move configuration of public addresses into a config persistent database
291 - infrastructure for the above
292 * Fri Aug 20 2010 : Version 1.2
293 - Initial version 1.2 release
294 * Tue May 25 2010 : Version 1.9
296 * Wed Mar 25 2010 : Version 1.0.114
297 - Lots of changes from Metze
298 * Wed Jan 13 2010 : Version 1.0.113
299 - Incorrect use of dup2() could cause ctdb to spin eating 100% cpu.
300 * Tue Jan 12 2010 : Version 1.0.112
301 - Revert the use of wbinfo --ping-dc as it is proving too unreliable.
302 - Minor testsuite changes.
303 * Fri Dec 18 2009 : Version 1.0.111
304 - Fix a logging bug when an eventscript is aborted that could cause a crash.
305 - Add back cb_status that was lost in a previous commit.
306 * Fri Dec 18 2009 : Version 1.0.110
307 - Metxe: fix for filedescriptor leak in the new eventscript code.
308 - Rusty: fix for a crash bug in the eventscript code.
309 * Thu Dec 17 2009 : Version 1.0.109
310 - Massive eventscript updates. (bz58828)
311 - Nice the daemon instead of using realtime scheduler, also use mlockall() to
312 reduce the risk of blockign due to paging.
313 - Workarounds for valgrind when forking once for each script. Valgrind consumes
314 massive cpu when terminating the scripts on virtual systems.
315 - Sync the tdb library with upstream, and use the new TDB_DISALLOW_NESTING flag.
316 - Add new command "ctdb dumpdbbackup"
317 - Start using the new tdb check framework to validate tdb files upon startup.
318 - A new framework where we can control health for individual tdb databases.
319 - Fix a crash bug in the logging code.
320 - New transaction code for persistent databases.
321 - Various other smaller fixes.
322 * Mon Dec 7 2009 : Version 1.0.108
323 - Transaction updates from Michael Adam.
324 - Use the new wbinfo --ping-dc instead of -p in the eventscript for samba
325 the check if winbindd is ok.
326 - Add a better "process-exist" for samba so it will automatically
327 reap smbd's on stopped and banned nodes to reclaim subrecords.
328 This will be done a bit differently in the next release.
329 - Use a statically allocated buffer for the 'first-time' capture buffer
330 to reduce the pressure on malloc/free.
331 * Wed Dec 2 2009 : Version 1.0.107
332 - fix for rusty to solve a double-free that can happen when there are
333 multiple packets queued and the connection is destroyed before
334 all packets are processed.
335 * Tue Dec 1 2009 : Version 1.0.106
336 - Buildscript changes from Michael Adam
337 - Dont do a full recovery when there is a mismatch detected for ip addresses,
338 just do a less disruptive ip-reallocation
339 - When starting ctdbd, wait until all initial recoveries have finished
340 before we issue the "startup" event.
341 So dont start services or monitoring until the cluster has
343 - Major eventscript overhaul by Ronnie, Rusty and Martins and fixes of a few
345 * Thu Nov 19 2009 : Version 1.0.105
346 - Fix a bug where we could SEGV if multiple concurrent "ctdb eventscript ..."
347 are used and some of them block.
348 - Monitor the daemon from the syslog child process so we shutdown cleanly when
349 the main daemon terminates.
350 - Add a 500k line ringbuffer in memory where all log messages are stored.
351 - Add a "ctdb getlog <level>" command to pull log messages from the in memory
353 - From martin : fixes to cifs and nfs autotests
354 - from michael a : fix a bashism in 11.natgw
355 * Fri Nov 6 2009 : Version 1.0.104
356 - Suggestion from Metze, we can now use killtcp to kill local connections
357 for nfs so change the killtcp script to kill both directions of an NFS
359 We used to deliberately only kill one direction in these cases due to
361 - Suggestion from christian Ambach, when using natgw, try to avoid using a
362 UNHEALTHY node as the natgw master.
363 - From Michael Adam: Fix a SEGV bug in the recent change to the eventscripts
364 to allow the timeout to apply to each individual script.
365 - fix a talloc bug in teh vacuuming code that produced nasty valgrind
367 - From Rusty: Set up ulimit to create core files for ctdb, and spawned
368 processes by default. This is useful for debugging and testing but can be
369 disabled by setting CTDB_SUPRESS_COREFILE=yes in the sysconfig file.
370 - Remove the wbinfo -t check from the startup check that winbindd is happy.
371 - Enhance the test for bond devices so we also check if the sysadmin have
372 disabled all slave devices using "ifdown".
373 * Tue Nov 3 2009 : Version 1.0.103
374 - Dont use vacuuming on persistent databases
375 - Michael A : transaction updates to persistent databases
376 - Dont activate service automatically when installing the RPM. Leave this to the admin.
377 - Create a child process to send all log messages to, to prevent a hung/slow syslogd
378 from blocking the main daemon. In this case, discard log messages instead and let the child
380 - Michael A: updates to log messages
381 * Thu Oct 29 2009 : Version 1.0.102
382 - Wolfgang: fix for the vacuuming code
383 - Wolfgang: stronger tests for persistent database filename tests
384 - Improve the log message when we refuse to startup since wbinfo -t fails
385 to make it easier to spot in the log.
386 - Update the uptime command output and the man page to indicate that
387 "time since last ..." if from either the last recovery OR the last failover
388 - Michael A: transaction updates
389 * Wed Oct 28 2009 : Version 1.0.101
390 - create a separate context for non-monitoring events so they dont interfere with the monitor event
391 - make sure to return status 0 in teh callback when we abort an event
392 * Wed Oct 28 2009 : Version 1.0.100
393 - Change eventscript handling to allow EventScriptTimeout for each individual script instead of for all scripts as a whole.
394 - Enhanced logging from the eventscripts, log the name and the duration for each script as it finishes.
395 - Add a check to use wbinfo -t for the startup event of samba
396 - TEMP: allow clients to attach to databases even when teh node is in recovery mode
397 - dont run the monitor event as frequently after an event has failed
398 - DEBUG: in the eventloops, check the local time and warn if the time changes backward or rapidly forward
399 - From Metze, fix a bug where recovery master becoming unhealthy did not trigger an ip failover.
400 - Disable the multipath script by default
401 - Automatically re-activate the reclock checking if the reclock file is specified at runtime. Update manpage to reflect this.
402 - Add a mechanism where samba can register a SRVID and if samba unexpectedly disconnects, a message will be broadcasted to all other samba daemons.
403 - Log the pstree on hung scripts to a file in /tmp isntead of /var/log/messages
404 - change ban count before unhealthy/banned to 10
405 * Thu Oct 22 2009 : Version 1.0.99
406 - Fix a SEGV in the new db priority code.
407 - From Wolfgang : eliminate a ctdb_fatal() if there is a dmaster violation detected.
408 - During testing we often add/delete eventscripts at runtime. This could cause an eventscript to fail and mark the node unhealthy if an eventscript was deleted while we were listing the names. Handle the errorcode and make sure the node does not becomne unhealthy in this case.
409 - Lower the debuglevel for the messages when ctdb creates a filedescruiptor so we dont spam the logs with these messages.
410 - Dont have the RPM automatically restart ctdb
411 - Volker : add a missing transaction_cancel() in the handling of persistent databases
412 - Treat interfaces with the anme ethX* as bond devices in 10.interfaces so we do the correct test for if they are up or not.
413 * Tue Oct 20 2009 : Version 1.0.98
414 - Fix for the vacuuming database from Wolfgang M
415 - Create a directory where the test framework can put temporary overrides
416 to variables and functions.
417 - Wait a lot longer before shutting down the node when the reclock file
418 is incorrectly configured, and log where it is configured.
419 - Try to avoid running the "monitor" event when databases are frozen.
420 - Add logging for every time we create a filedescriptor so we can trap
422 * Thu Oct 14 2009 : Version 1.0.97
423 - From martins : update onnode.
424 Update onnode to allow specifying an alternative nodes file from
425 the command line and also to be able to specify hostnames on the
427 onnode host1,host2,...
428 * Tue Oct 13 2009 : Version 1.0.96
429 - Add more debugging output when eventscripts have trouble. Print a
430 "pstree -p" to the log when scripts have hung.
431 - Update the initscript, only print the "No reclock file used" warning
432 when we do "service ctdb start", dont also print them for all other
434 - When changing between unhealthy/healthy state, push a request to the
435 recovery master to perform an ip reallocation instead of waiting for the
436 recovery master to pull and check the state change.
437 - Fix a bug in the new db-priority handling where a pre-.95 recovery master
438 could no longer lock the databases on a post-.95 daemon.
439 - Always create the nfs state directories during the "monitor" event.
440 This makes it easier to configure and enable nfs at runtime.
441 - From Volker, forward-port a simper deadlock avoiding patch from the 1.0.82
442 branch. This is a simpler versionof the "db priority lock order" patch
443 that went into 1.0.95, and will be kept for a few versions until samba
444 has been updated to use the functionality from 1.0.95.
445 * Mon Oct 12 2009 : Version 1.0.95
446 - Add database priorities. Allow samba to set the priority of databases
447 and lock the databases in priority order during recovery
448 to avoid a deadlock when samba locks one database then blocks indefinitely
449 while waiting for the second databaso to become locked.
450 - Be aggressive and ban nodes where the recovery transaction start call
452 * Thu Oct 10 2009 : Version 1.0.94
453 - Be very aggressive and quickly ban nodes that can not freeze their databases
454 * Tue Oct 8 2009 : Version 1.0.93
455 - When adding an ip, make sure to update this assignment on all nodes
456 so it wont show up as -1 on other nodes.
457 - When adding an ip and immediately deleting it, it was possible that
458 the daemon would crash accessing already freed memory.
459 Readjust the memory hierarchy so the destructors are called in the right order.
460 - Add a handshake to the recovery daemon to eliminate some rare cases where
461 addip/delip might cause a recovery to occur.
462 - updated onnode documenation from Martin S
463 - Updates to the natgw eventscript to allow disabling natgw at runtime
464 * Fri Oct 2 2009 : Version 1.0.92
465 - Test updates and merge from martin
466 - Add notification for "startup"
467 - Add documentation for notification
468 - from martin, a fix for restarting vsftpd in the eventscript
469 * Tue Sep 29 2009 : Version 1.0.91
470 - New vacuum and repack design from Wolgang Mueller.
471 - Add a new eventscript 01.reclock that will first mark a node unhealthy and later ban the node if the reclock file can not be accessed.
472 - Add machinereadable output to the ctdb getreclock command
473 - merge transaction updates from Michael Adam
474 - In the new banning code, reset the culprit count to 0 for all nodes that could successfully compelte a full recovery.
475 - dont mark the recovery master as a ban culprit because a node in the cluster needs a recovery. this happens naturally when using ctdb recover command so dont make this cause a node to be banned.
476 * Sat Sep 12 2009 : Version 1.0.90
477 - Be more forgiving for eventscripts that hang during startup
478 - Fix for a banning bug in the new banning logic
479 * Thu Sep 3 2009 : Version 1.0.89
480 - Make it possible to manage winbind independently of samba.
481 - Add new prototype banning code
482 - Overwrite the vsftpd state file instead of appending. This eliminates
483 annoying errors in the log.
484 - Redirect some iptables commands to dev null
485 - From Michael A, explicitely set the broadcast when we takeover a public ip
486 - Remove a reclock file check we no longer need
487 - Skip any persistent database files ending in .bak
488 * Mon Aug 17 2009 : Version 1.0.88
489 - Add a new state for eventscripts : DISABLED.
490 Add two new commands "ctdb enablescript/disablescript" to enable/disable
491 eventscripts at runtime.
492 - Bugfixes for TDB from rusty.
493 - Merge/Port changes from upstream TDB library by rusty.
494 - Additional new tests from MartinS. Tests for stop/continue.
495 - Initial patch to rework vacuuming/repacking process from Wolfgang Mueller.
496 - Updates from Michael Adam for persistent writes.
497 - Updates from MartinS to handle the new STOPPED bit in the test framework.
498 - Make it possible to enable/disable the RECMASTER and LMASTER roles
499 at runtime. Add two new commands
500 "ctdb setlmasterrole/setrecmasterrole on/off"
501 - Make it possible to enable/disable the natgw feature at runtime. Add
502 the command "ctdb setnatgwstate on/off"
503 * Fri Jul 17 2009 : Version 1.0.87
504 - Add a new event "stopped" that is called when a node is stopped.
505 - Documentation of the STOPPED flag and the stop/continue commands
506 - Make it possible to start a node in STOPPED mode.
507 - Add a new node flag : STOPPED and commands "ctdb stop" "ctdb continue"
508 These commands are similar to "diasble/enable" but will also remove the node from the vnnmap, while disable only fails all ip addresses over.
509 - tests for NFS , CIFS by martins
510 - major updates to the init script by martins
511 - Send gratious arps with a 1.1 second stride instead of a 1 second stride to workaround interesting "features" of common linux stacks.
512 - Various test enhancements from martins:
513 - additional other tests
514 - add tests for grat arp generation, ping during failover, ssh and failover
515 - New/updated tcp tickle tests and supprot functions
516 - provide better debugging when a test fails
517 - make ctdbd restarts more reliable in the tests
518 - update the "wait bar" to make the wait progress in tests more obvious
520 - when dispatching a message to a handler, make the message a real talloc object so that we can reparent the object in the tallic hierarchy.
521 - document the ipreallocate command
522 - Updates to enable/disable to use the ipreallocate command to block until the following ipreallocation has completed.
523 - Update the main daemon and the tools to allow debug level to be a string instead of an integer.
524 - Update the sysconfig file to show using string literals instead of numeric values for the debuglevels used.
525 - If no debuglevel is specific, make "ctdb setdebug" show the available options.
526 - When trying to allocate network packets, add explicit checks if the network transport has been shutdown before trying and failing, to make log messages easier to read. Add this extra check and logging to every plave packets are allocated.
527 * Tue Jun 30 2009 : Version 1.0.86
528 - Do not access the reclock at all if VerifyRecoveryLock is zero, not even try to probe it.
529 - Allow setting the reclock file as "", which means that no reclock file at all should be used.
530 - Document that a reclock file is no longer required, but that it is dangerous.
531 - Add a control that can be used to set/clear/change the reclock file in the daemon during runtime.
532 - Update the recovery daemon to poll whether a reclock file should be sued and if so which file at runtime in each monitoring cycle.
533 - Automatically disable VerifyRecoveryLock everytime a user changes the location of the reclock file.
534 - do not allow the VerifyRecoveryLock to be set using ctdb setvar if there is no recovery lock file specified.
535 - Add two commands "ctdb getreclock" and "ctdb setreclock" to modify the reclock file.
536 * Tue Jun 23 2009 : Version 1.0.85
537 - From William Jojo : Dont use getopt on AIX
538 - Make it possible to use "ctdb listnodes" also when the daemon is not running
539 - Provide machinereadable output to "ctdb listnodes"
540 - Dont list DELETED nodes in the ctdb listnodes output
541 - Try to avoid causing a recovery for the average case when adding/deleting/moving an ip
542 - When banning a node, drop the IPs on that node only and not all nodes.
543 - Add tests for NFS and CIFS tickles
544 - Rename 99.routing to 11.routing so it executes before NFS and LVS scripts
545 - Increase the default timeout before we deem an unresponsive recovery daemon hung and shutdown
546 - Reduce the reclock timout to 5 seconds
547 - Spawn a child process in the recovery daemon ot check the reclock file to
548 avoid blocking the process if the underlying filesystem is unresponsive
549 - fix for filedescriptor leak when a child process timesout
550 - Dont log errors if waitpid() returns -1
551 - Onnode updates by Martins
552 - Test and initscript cleanups from Martin S
553 * Tue Jun 2 2009 : Version 1.0.84
554 - Fix a bug in onnode that could not handle dead nodes
555 * Tue Jun 2 2009 : Version 1.0.83
556 - Document how to remove a ndoe from a running cluster.
557 - Hide all deleted nodes from ctdb output.
558 - Lower the loglevel on some eventscript related items
559 - Dont queue packets to deleted nodes
560 - When building initial vnnmap, ignode any nonexisting nodes
561 - Add a new nodestate : DELETED that is used when deleting a node from an
563 - dont remove the ctdb socket when shutting down. This prevents a race in the
564 initscripts when restarting ctdb quickly after stopping it.
565 - TDB nesting reworked.
566 - Remove obsolete ipmux
567 - From Flavio Carmo Junior: Add eventscript and documentation for ClamAV antivirus engine
568 - From Sumit Bose: fix the regex in the test to handle the new ctdb
569 statistics output that was recently added.
570 - change the socket type we use for grauitious arps from the obsolete
571 AF_INET/SOCK_PACKET to instead use PF_PACKET/SOCK_RAW.
572 - Check return codes for some functions, from Sumit Bose, based on codereview by Jim Meyering.
573 - Sumit Bose: Remove structure memeber node_list_file that is no longer used.
574 - Sumit Bose: fix configure warning for netfilter.h
575 - Updates to the webpages by Volker.
576 - Remove error messages about missing /var/log/log.ctdb file from ctdb_diagnostics.sh from christian Ambach
577 - Additional error logs if hte eventscript switching from dameon to client mode fails.
578 - track how long it takes for ctdbd and the recovery daemon to perform the rec-lock fcntl() lock attemt and show this in the ctdb statistics output.
579 * Thu May 14 2009 : Version 1.0.82
580 - Update the "ctdb lvsmaster" command to return -1 on error.
581 - Add a -Y flag to "ctdb lvsmaster"
582 - RHEL5 apache leaks semaphores when crashing. Add semaphore cleanup to the
583 41.httpd eventscript and try to restart apache when it has crashed.
584 - Fixes to some tests
585 - Add a -o option to "onnode" which will redirect all stdout to a file for
587 - Add a natgw and a lvs node specifier to onnode so that we can use
589 - Assign the natgw address to lo instead of the private network so it can also
590 be used where private and public networks are the same.
591 - Add GPL boilerplates to two missing scripts.
592 - Change the natgw prefix NATGW_ to CTDB_NATGW_
593 * Fri May 8 2009 : Version 1.0.81
594 - use smbstatus -np instead of smbstatus -n in the 50.samba eventscript
595 since this avoids performing an expensive traverse on the locking and brlock
597 - make ctdb automatically terminate all traverse child processes clusterwide
598 associated to a client application that terminates before the traversal is
600 - From Sumit Bose : fixes to AC_INIT handling.
601 - From Michael Adam, add Tridge's "ping_pong" tool the the ctdb distro since
602 this is very useful for testing the backend filesystem.
603 - From Sumit bose, add support for additional 64 bit platforms.
604 - Add a link from the webpage to Michael Adams SambaXP paper on CTDB.
605 * Fri May 1 2009 : Version 1.0.80
606 - change init shutdown level to 01 for ctdb so it stops before any of the other services
607 - if we can not pull a database from a remote node during recovery, mark that node as a culprit so it becomes banned
608 - increase the loglevel when we volunteer to drop all ip addresses after beeing in recovery mode for too long. Make this timeout tuneable with "RecoveryDropAllIPs" and have it default to 60 seconds
609 - Add a new flag TDB_NO_NESTING to the tdb layer to prevent nested transactions which ctdb does not use and does not expect. Have ctdb set this flag to prevent nested transactions from occuring.
610 - dont unconditionally kill off ctdb and restrat it on "service ctdb start". Fail "service ctdb start" with an error if ctdb is already running.
611 - Add a new tunable "VerifyRecoveryLock" that can be set to 0 to prevent the main ctdb daemon to verify that the recovery master has locked the reclock file correctly before allowing it to set the recovery mode to active.
612 - fix a cosmetic bug with ctdb statistics where certain counters could become negative.
613 * Wed Apr 8 2009 : Version 1.0.79
614 - From Mathieu Parent: add a ctdb pkgconfig file
616 - add a funciton remove_ip to safely remove an ip from an interface, taking care to workaround an issue with linux alias interfaces.
617 - Update the natgw eventscript to use the safe remove_ip() function
618 - fix a bug in the eventscript child process that would cause the socket to be removed.
619 - dont verify nodemap on banned nodes during cluster monitoring
620 - Update the dodgy SeqnumInterval to have ms resolution
621 * Tue Mar 31 2009 : Version 1.0.78
622 - Add a notify mechanism so we can send snmptraps/email to external management systems when the node becomes unhealthy
623 - include 11.natgw eventscript in thew install so that the NATGW feature works
624 * Tue Mar 31 2009 : Version 1.0.77
625 - Update the 99.routing eventscript to also try to add the routes (back) during a releaseip event. Similar to the reasons why we must add addresses back during releaseip in 10.interfaces
626 * Wed Mar 24 2009 : Version 1.0.76
627 - Add a debugging command "xpnn" which can print the pnn of the node even when ctdbd is not running.
628 - Redo the NATGW implementation to allow multiple disjoing NATGW groups in the same cluster.
629 * Tue Mar 24 2009 : Version 1.0.75
630 - Various updates to LVS
631 - Fix a bug in the killtcp control where we did not set the port correctly
632 - add a new "ctdb scriptstatus" command that shows the status of the eventrscripts.
633 * Mon Mar 16 2009 : Version 1.0.74
634 - Fixes to AIX from C Cowan.
635 - Fixes to ctdb_diagnostics so we collect correct GPFS data
636 - Fixes to the net conf list command in ctdb_diagnostics
637 - Check the static-routes file IFF it exists in ctdb_diagnostics
638 * Wed Mar 4 2009 : Version 1.0.73
639 - Add possibility to disable the check of shares for NFS and Samba
640 - From Sumit Bose, fix dependencies so make -j works
641 * Wed Feb 18 2009 : Version 1.0.72
642 - Updates to test scripts by martin s
643 - Adding a COPYING file
644 - Use netstat to check for services and ports and fallback to netcat
645 only if netstat is unavailable.
646 * Mon Feb 2 2009 : Version 1.0.71
647 - Additional ipv6 fixes from Michael Adams
648 * Fri Jan 16 2009 : Version 1.0.70
649 - IPv6 support is completed. this is backward compatible with ipv4-only
650 systems. To use IPv6 with samba and ctdb you need current GIT of samba 3.3
651 or michael adams samba-ctdeb branch.
652 - Many enhancements to the build system and scripts to make it more SUSE
653 friendly by Michael Adams.
654 - Change of how the naming of the package is structured. We are now
655 using "1.0.70" as a release and "-1" as the revision instead of as
656 previously using "1.0" as release and ".70" as the revision.
658 * Thu Dec 18 2008 : Version 1.0.69
659 - Various fixes to scripts by M Adam
660 - Dont call ctdb_fatal() when the transport is down during shutdown
661 * Fri Dec 12 2008 : Version 1.0.68
662 - Fixes for monitoring of interfaces status from Michael Adam.
663 - Use -q instead of >/dev/null for grep to enhance readability of the
664 scripts from Michael Adam.
665 - Update to the "ctdb recover" command. This command now block until the
666 has completed. This makes it much easier to use in scripts and avoids
667 the common workaround :
669 ... loop while waiting for recovery completes ...
671 - Add a CTDB_TIMEOUT variable. If set, this variable provides an automatic
672 timeout for "ctdb <command>", similar to using -T <timeout>
673 - Set a unique errorcode for "ctdb <command>" when it terminates due to a
674 timeout so that scripts can distinguish between a hung command and what was
676 - Update "ctdb ban/unban" so that if the cluster is in recovery these commands
677 blocks and waits until after recovery is complete before the perform the
678 ban/unban operation. This is necessary since the recovery process can cause
679 nodes to become automatically unbanned.
680 - Update "ctdb ban/unban" to block until the recovery that will follow shortly
681 after this command has completed.
682 This makes it much easier to use in scripts and avoids the common
685 ... loop while waiting for recovery completes ...
687 - Bugfix for the new flags handling in 1.0.67. Abort and restart monitoring
688 if we failed to get proper nodemaps from a remote node instead of
689 dereferencing a null pointer.
690 - If ctdbd was explicitely started with the '--socket' argument, make
691 ctdbd automatically set CTDB_SOCKET to the specified argument.
692 This ensures that eventscripts spawned by the ctdb daemon will default to
693 using the same socket and talk to the correct daemon.
694 This primarily affects running multiple daemons on the same host and where
695 you want each instance of ctdb daemons have their eventscripts talk to the
697 - Update "ctdb ping" to return an error code if the ping fail so that it
698 can be used in scripts.
699 - Update to how to synchronize management of node flags across the cluster.
700 * Thu Dec 3 2008 : Version 1.0.67
701 - Add a document describing the recovery process.
702 - Fix a bug in "ctdb setdebug" where it would refuse to set a negative
704 - Print the list of literals for debug names if an invalid one was given
706 - Redesign how "ctdb reloadnodes" works and reduce the amont of tcp teardowns
707 used during this event.
708 - Make it possible to delete a public ip from all nodes at once using
710 * Mon Nov 24 2008 : Version 1.0.66
711 - Allow to change the recmaster even when we are not frozen.
712 - Remove two redundant SAMBA_CHECK variables from the sysconf example
713 - After a node failure it can take very long before some lock operations
714 ctdb needs to perform are allowed/works with gpfs again. Workaround this
715 by treating a hang/timeout as success.
716 - Dont override CTDB_BASE is fet in the shell already
717 - Always send keepalive packets regardless of whether the link is idle or not.
718 - Rewrite the disable/enable flag update logic to prevent a race between
719 "ctdb disable/enable" and the recovery daemon when updating the flags to
721 * Thu Nov 13 2008 : Version 1.0.65
722 - Update the sysconfig example: The default debug level is 2 (NOTICE) and not
724 - Add support for a CTDB_SOCKET environment variable for the ctdb command
725 line tool. If set, this overrides the default socket the ctdb tool will
727 - Add logging of high latency operations.
728 * Mon Oct 22 2008 : Version 1.0.64
729 - Add a context and a timed event so that once we have been in recovery for
730 too long we drop all public addresses.
731 * Mon Oct 20 2008 : Version 1.0.63
732 - Remove logging of "periodic cleanup ..." in 50.samba
733 - When we reload a nodes file, we must detect this and reload the file also
734 in the recovery daemon before we try to dereference somethoung beyond the end
736 * Thu Oct 16 2008 : Version 1.0.62
737 - Allow multiple eventscritps using the same prefix number.
738 It is undefined which order scripts with the same prefix will execute in.
739 * Wed Oct 15 2008 : Version 1.0.61
740 - Use "route add -net" instead of "ip route add" when adding routes in 99.routing
741 - lower the loglevel os several debug statements
742 - check the status returned from ctdb_ctrl_get_tickles() before we try to print them out to the screen.
743 - install a new eventscript 20.multipathd whoich can be used to monitor that multipath devices are healthy
744 * Wed Oct 15 2008 : Version 1.0.60
745 - Verify that nodes we try to ban/unban are reachable and print an error othervise.
746 - Update the client and server sides of TAKEIP/RELEASEIP/GETPUBLICIPS and GETNODEMAP to fall back to the old style ipv4-only controls if the new ipv4/ipv6 controls fail. This allows an ipv4/v6 enabled ctdb daemon to interoperate with earlier ipv4-only versions of the daemons.
747 - From Mathieu Parent : log debian systems log the package versions in ctdb diagnostics
748 - From Mathieu Parent : specify logdir location for debian (this patch was later reversed)
749 - From Michael Adams : allow # comments in nodes/public_addresses files
750 * Tue Oct 7 2008 : Version 1.0.59
751 - Updated "reloadnodes" logic. Instead of bouncing the entire tcp layer it is sufficient to just close and reopen all outgoing tcp connections.
752 - New eventscript 99.routing which can be used to re-attach routes to public interfaces after a takeip event. (routes may be deleted by the kernel when we release an ip)
753 - IDR tree fix from Jim Houston
754 - Better handling of critical events if the local clock is suddenly changed forward by a lot.
755 - Fix three slow memory leaks in the recovery daemon
756 - New ctdb command : ctdb recmaster which prints the pnn of the recmaster
757 - Onnode enhancements from Martin S : "healthy" and "connected" are now possible nodespecifiers
758 - From Martin S : doc fixes
759 - lowering some debug levels for some nonvital informational messages
760 - Make the daemon daemon monitoring stronger and allow ctdbd to detect a hung
762 - From C Cowan : patches to compile ipv6 under AIX
763 - zero out some structs to keep valgrind happy
764 * Wed Aug 27 2008 : Version 1.0.58
765 - revert the name change tcp_tcp_client back to tcp_control_tcp so
767 - Updates to the init script from Abhijith Das <adas@redhat.com>
768 * Mon Aug 25 2008 : Version 1.0.57
769 - initial support for IPv6
770 * Mon Aug 11 2008 : Version 1.0.56
771 - fix a memory leak in the recovery daemon.
772 * Mon Aug 11 2008 : Version 1.0.55
773 - Fix the releaseip message we seond to samba.
774 * Fri Aug 8 2008 : Version 1.0.54
775 - fix a looping error in the transaction code
776 - provide a more detailed error code for persistent store errors
777 so clients can make more intelligent choices on how to try to recover
778 * Thu Aug 7 2008 : Version 1.0.53
779 - Remove the reclock.pnn file it can cause gpfs to fail to umount
780 - New transaction code
781 * Mon Aug 4 2008 : Version 1.0.52
782 - Send an explicit gratious arp when starting sending the tcp tickles.
783 - When doing failover, issue a killtcp to non-NFS/non-CIFS clients
784 so that they fail quickly. NFS and CIFS already fail and recover
786 - Update the test scripts to handle CTRL-C to kill off the test.
787 * Mon Jul 28 2008 : Version 1.0.51
788 - Strip off the vlan tag from bond devices before we check in /proc
789 if the interface is up or not.
790 - Use testparm in the background in the scripts to allow probing
791 that the shares do exist.
792 - Fix a bug in the logging code to handle multiline entries better
793 - Rename private elements from private to private_data
794 * Fri Jul 18 2008 : Version 1.0.50
795 - Dont assume that just because we can establish a TCP connection
796 that we are actually talking to a functioning ctdb daemon.
797 So dont mark the node as CONNECTED just because the tcp handshake
799 - Dont try to set the recmaster to ourself during elections for those
800 cases we know this will fail. To remove some annoying benign but scary
801 looking entries from the log.
802 - Bugfix for eventsystem for signal handling that could cause a node to
804 * Thu Jul 17 2008 : Version 1.0.49
805 - Update the safe persistent update fix to work with unpatched samba
807 * Thu Jul 17 2008 : Version 1.0.48
808 - Update the spec file.
809 - Do not start new user-triggered eventscripts if we are already
810 inside recovery mode.
811 - Add two new controls to start/cancel a persistent update.
812 A client such as samba can use these to tell ctdbd that it will soon
813 be writing directly to the persistent database tdb file. So if
814 samba is -9ed before it has eitehr done the persistent_store or
815 canceled the operation, ctdb knows that the persistent databases
816 'may' be out of sync and therefore a full blown recovery is called for.
817 - Add two new options :
818 CTDB_SAMBA_SKIP_CONF_CHECK and CTDB_SAMBA_CHECK_PORTS that can be used
819 to override what checks to do when monitoring samba health.
820 We can no longer use the smbstatus, net or testparm commands to check
821 if samba or its config is healthy since these commands may block
822 indefinitely and thus can not be used in scripts.
823 * Fri Jul 11 2008 : Version 1.0.47
824 - Fix a double free bug where if a user striggered (ctdb eventscript)
825 hung and while the timeout handler was being processed a new user
826 triggered eventscript was started we would free state twice.
827 - Rewrite of onnode and associated documentation.
828 * Thu Jul 10 2008 : Version 1.0.46
829 - Document both the LVS:cingle-ip-address and the REMOTE-NODE:wan-accelerator
831 - Add commands "ctdb pnn", "ctdb lvs", "ctdb lvsmaster".
832 - LVS improvements. LVS is the single-ip-address mode for a ctdb cluster.
833 - Fixes to supress rpmlint warnings
835 - Change \s to [[:space:]] in some scripts. Not all RHEL5 packages come
836 with a egrep that handles \s even same version but different arch.
837 - Revert the change to NFS restart. CTDB should NOT attempt to restart
839 - Rewrite of the waitpid() patch to use the eventsystem for handling
841 * Tue Jul 8 2008 : Version 1.0.45
842 - Try to restart the nfs service if it has failed to respond 3 times in a row.
843 - waitpid() can block if the child does not respond promptly to SIGTERM.
844 ignore all SIGCHILD signals by setting SIGCHLD to SIG_DEF.
845 get rid of all calls to waitpid().
846 - make handling of eventscripts hanging more liberal.
847 only consider the script to have failed and making the node unhealthy
848 IF the eventscript terminated wiht an error
849 OR the eventscript hung 5 or more times in a row
850 * Mon Jul 7 2008 : Version 1.0.44
851 - Add a CTDB_VALGRIND option to /etc/sysconfig/ctdb to make it start
852 ctdb under valgrind. Logs go to /var/log/ctdb_valgrind.PID
853 - Add a hack to show the control opcode that caused uninitialized data
854 in the valgrind output by encoding the opcode as the line number.
855 - Initialize structures and allocated memory in various places in
856 ctdb to make it valgrind-clean and remove all valgrind errors/warnings.
857 - If/when we destroy a lockwait child, also make sure we cancel any pending transactions
858 - If a transaction_commit fails, delete/cancel any pending transactions and
859 return an error instead of calling ctdb_fatal()
860 - When running ctdb under valgrind, make sure we run it with --nosetsched and also
861 ensure that we do not use mem-mapped i/o when accessing the tdb's.
862 - zero out ctdb->freeze_handle when we free/destroy a freeze-child.
863 This prevent a heap corruption/ctdb crash bug that could trigger
864 if the freeze child times out.
865 - we dont need to explicitely thaw the databases from the recovery daemon
866 since this is done implicitely when we restore the recovery mode back to normal.
867 - track when we start and stop a recovery. Add the 'time it took to complete the
868 recovery' to the 'ctdb uptime' output.
869 Ensure by tracking the start/stop recovery timestamps that we do not
870 check that the ip allocation is consistend from inside the recovery daemon
871 while a different node (recovery master) is performing a recovery.
872 This prevent a race that could cause a full recovery to trigger if the
873 'ctdb disable/enable' commands took very long.
874 - The freeze child indicates to the master daemon that all databases are locked
875 by writing data to the pipe shared with the master daemon.
876 This write sometimes fail and thus the master daemon never notices that the databases
877 are locked cvausing long timeouts and extra recoveries.
878 Check that the write is successful and try the write again if it failed.
879 - In each node, verify that the recmaster have the right node flags for us
880 and force a push of our flags to the recmaster if wrong.
881 * Tue Jul 1 2008 : Version 1.0.43
882 - Updates and bugfixes to the specfile to keep rpmlint happy
883 - Force a global flags update after each recovery event.
884 - Verify that the recmaster agrees with our node flags and update the
886 - When writing back to the parent from a freeze-child across the pipe,
887 loop over the write in case the write failed with an error othervise
888 the parent will never be notified tha the child has completed the operation.
889 - Automatically thaw all databases when recmaster marks us as being in normal
890 mode instead of recovery mode.
891 * Fri Jun 13 2008 : Version 1.0.42
892 - When event scripts have hung/timedout more than EventScriptBanCount times
893 in a row the node will ban itself.
894 - Many updates to persistent write tests and the test scripts.
895 * Wed May 28 2008 : Version 1.0.41
896 - Reactivate the safe writes to persistent databases and solve the
897 locking issues. Locking issues are solved the only possible way,
898 by using a child process to do the writes. Expensive and slow but... .
899 * Tue May 27 2008 : Version 1.0.40
900 - Read the samba sysconfig file from the 50.samba eventscript
901 - Fix some emmory hierarchical bugs in the persistent write handling
902 * Thu May 22 2008 : Version 1.0.39
903 - Moved a CTDB_MANAGES_NFS, CTDB_MANAGES_ISCSI and CTDB_MANAGES_CSFTPD
904 into /etc/sysconfig/ctdb
905 - Lowered some debug messages to not fill the logfile with entries
906 that normally occur in the default configuration.
907 * Fri May 16 2008 : Version 1.0.38
908 - Add machine readable output support to "ctdb getmonmode"
909 - Lots of tweaks and enhancements if the event scripts are "slow"
910 - Merge from tridge: an attempt to break the chicken-and-egg deadlock that
911 net conf introduces if used from an eventscript.
912 - Enhance tickles so we can tickle an ipv6 connection.
913 - Start adding ipv6 support : create a new container to replace sockaddr_in.
914 - Add a checksum routine for ipv6/tcp
915 - When starting up ctdb, let the init script do a tdbdump on all
916 persistent databases and verify that they are good (i.e. not corrupted).
917 - Try to use "safe transactions" when writing to a persistent database
918 that was opened with the TDB_NOSYNC flag. If we can get the transaction
919 thats great, if we cant we have to write anyway since we cant block here.
920 * Mon May 12 2008 : Version 1.0.37
921 - When we shutdown ctdb we close the transport down before we run the
922 "shutdown" eventscripts. If ctdb decides to send a packet to a remote node
923 after we have shutdown the transport but before we have shutdown ctdbd
924 itself this could lead to a SEGV instead of a clean shutdown. Fix.
925 - When using the "exportfs" command to extract which NFS export directories
926 to monitor, exportfs violates the "principle of least surprise" and
927 sometimes report a single export line as two lines of text output
928 causing the monitoring to fail.
929 * Fri May 9 2008 : Version 1.0.36
930 - fix a memory corruption bug that could cause the recovery daemon to crash.
931 - fix a bug with distributing public ip addresses during recovery.
932 If the node that is the recovery master did NOT use public addresses,
933 then it assumed that no other node in the cluster used them either and
934 thus skipped the entire step of reallocating public addresses.
935 * Wed May 7 2008 : Version 1.0.35
936 - During recovery, when we define the new set of lmasters (vnnmap)
937 only consider those nodes that have the can-be-lmaster capability
938 when we create the vnnmap. unless there are no nodes available which
939 supports this capability in which case we allow the recmaster to
940 become lmaster capable (temporarily).
941 - Extend the async framework so that we can use paralell async calls
942 to controls that return data.
943 - If we do not have the "can be recmaster" capability, make sure we will
944 lose any recmaster elections, unless there are no nodes available that
945 have the capability, in which case we "take/win" the election anyway.
946 - Close and reopen the reclock pnn file at regular intervals.
947 Make it a non-fatal event if we occasionally fail to open/read/write
949 - Monitor that the recovery daemon is still running from the main ctdb
950 daemon and shutdown the main daemon when recovery daemon has terminated.
951 - Add a "ctdb getcapabilities" command to read the capabilities off a node.
952 - Define two new capabilities : can be recmaster and can be lmaster
953 and default both capabilities to YES.
954 - Log denied tcp connection attempts with DEBUG_ERR and not DEBUG_WARNING
955 * Thu Apr 24 2008 : Version 1.0.34
956 - When deleting a public ip from a node, try to migrate the ip to a different
958 - Change catdb to produce output similar to tdbdump
959 - When adding a new public ip address, if this ip does not exist yet in
960 the cluster, then grab the ip on the local node and activate it.
961 - When a node disagrees with the recmaster on WHO is the recmaster, then
962 mark that node as a recovery culprit so it will eventually become
964 - Make ctdb eventscript support the -n all argument.
965 * Thu Apr 10 2008 : Version 1.0.33
966 - Add facilities to include site local adaptations to the eventscript
967 by /etc/ctdb/rc.local which will be read by all eventscripts.
968 - Add a "ctdb version" command.
969 - Secure the domain socket with proper permissions from Chris Cowan
970 - Bugfixes for AIX from Chris Cowan
971 * Wed Apr 02 2008 : Version 1.0.32
972 - Add a control to have a node execute the eventscripts with arbitrary
973 command line arguments.
974 - Add a control "rddumpmemory" that will dump the talloc memory allocations
975 for the recovery daemon.
976 - Decorate the talloc memdump to produce better and easier memory leak
978 - Update the RHEL5 iscsi tgtd scripts to allow one iscsi target for each
980 - Add two new controls "addip/delip" that can be used to add/remove public
981 addresses to a node at runtime. After using these controls a "ctdb recover"
982 ir required to make the changes take.
983 - Fix a couple of slow memory leaks.
984 * Tue Mar 25 2008 : Version 1.0.31
985 - Add back controls to disable/enable monitoring on a node.
986 - Fix a memory leak where we used to attach CALL data to the ctdb structure
987 when performing a local call. Memory which would be lost if the call was
989 - Reduce the loglevel for the log output when someone connects to a non
990 public ip address for samba.
991 - Redo and optimize the vacuuming process to send only one control to each
992 other node containing all records to be vacuumed instead of one
993 control per node per record.
994 * Tue Mar 04 2008 : Version 1.0.30
995 - Update documentation cor new commands and tuneables
996 - Add machinereadable output to the ip,uptime and getdebug commands
997 - Add a moveip command to manually failover/failback public ips
998 - Add NoIPFallback tuneable that prevents ip address failback
999 - Use file locking inside the CFS as alternative to verify when other nodes
1000 Are connected/disconnected to be able to recover from split network
1001 - Add DisableWhenUnhealthy tunable
1002 - Add CTDB_START_AS_DISABLED sysconfig param
1003 - Add --start-as-disabled flag to ctdb
1004 - Add ability to monitor for OOM condition
1005 * Thu Feb 21 2008 : Version 1.0.29
1006 - Add a new command to make expansion of an existing cluster easier
1007 - Fix bug with references to freed objects in the ctdb structure
1008 - Propagate debuglevel changes to the recovery daemon
1009 - Merge patches to event scripts from Mathieu Parent :
1010 - MP: Simulate "service" on systems which do not provide this tool
1011 - MP: Set correct permissions for events.d/README
1012 - Add nice helper functions to start/stop nfs from the event scripts
1013 * Fri Feb 08 2008 : Version 1.0.28
1014 - Fix a problem where we tried to use ethtool on non-ethernet interfaces
1015 - Warn if the ipvsadm packege is missing when LVS is used
1016 - Dont use absolute pathnames in some of the event scripts
1017 - Fix for persistent tdbs growing inifinitely.
1018 * Wed Feb 06 2008 : Version 1.0.27
1019 - Add eventscript for iscsi
1020 * Thu Jan 31 2008 : Version 1.0.26
1021 - Fix crashbug in tdb transaction code
1022 * Tue Jan 29 2008 : Version 1.0.25
1023 - added async recovery code
1024 - make event scripts more portable
1025 - fixed ctdb dumpmemory
1026 - more efficient tdb allocation code
1027 - improved machine readable ctdb status output
1029 * Wed Jan 16 2008 : Version 1.0.24
1030 - added syslog support
1031 - documentation updates
1032 * Wed Jan 16 2008 : Version 1.0.23
1033 - fixed a memory leak in the recoveryd
1034 - fixed a corruption bug in the new transaction code
1035 - fixed a case where an packet for a disconnected client could be processed
1036 - added http event script
1037 - updated documentation
1038 * Thu Jan 10 2008 : Version 1.0.22
1039 - auto-run vacuum and repack ops
1040 * Wed Jan 09 2008 : Version 1.0.21
1041 - added ctdb vacuum and ctdb repack code
1042 * Sun Jan 06 2008 : Version 1.0.20
1043 - new transaction based recovery code
1044 * Sat Jan 05 2008 : Version 1.0.19
1045 - fixed non-master bug
1046 - big speedup in recovery for large databases
1047 - lots of changes to improve tdb and ctdb for high churn databases
1048 * Thu Dec 27 2007 : Version 1.0.18
1049 - fixed crash bug in monitor_handler
1050 * Tue Dec 04 2007 : Version 1.0.17
1051 - fixed bugs related to ban/unban of nodes
1052 - fixed a race condition that could lead to monitoring being permanently disabled,
1053 which would lead to long recovery times
1054 - make deterministic IPs the default
1055 - fixed a bug related to continuous recovery
1056 - added a debugging option --node-ip