1 %define with_systemd
%{?_with_systemd: 1} %{?!_with_systemd: 0}
2 %define initdir
%{_sysconfdir}/init.d
6 Packager: Samba Team
<samba@samba.org
>
10 License: GNU GPL version
3
11 Group: System Environment
/Daemons
12 URL: http
://ctdb.samba.org
/
14 Source: ctdb
-%{version}.tar.gz
17 Requires: coreutils
, sed
, gawk
, iptables
, iproute
, procps
, ethtool
, sudo
18 # Commands - package name might vary
19 Requires: /usr
/bin
/killall
, /bin
/kill, /bin
/netstat
21 Provides: ctdb
= %{version}
24 BuildRoot: %{_tmppath}/%{name}-%{version}-root
26 # Allow build with system libraries
27 # To enable, run rpmbuild with,
28 # "--with system_talloc"
30 # "--with system_tevent"
31 #%define with_included_talloc %{?_with_system_talloc: 0} %{?!_with_system_talloc: 1}
32 #%define with_included_tdb %{?_with_system_tdb: 0} %{?!_with_system_tdb: 1}
33 #%define with_included_tevent %{?_with_system_tevent: 0} %{?!_with_system_tevent: 1}
35 %define with_included_talloc
0
36 %define with_included_tevent
0
37 %define with_included_tdb
0
39 # Required minimum library versions when building with system libraries
40 %define libtalloc_version
2.0.8
41 %define libtdb_version
1.2.11
42 %define libtevent_version
0.9.16
44 %if ! %with_included_talloc
45 BuildRequires
: libtalloc
-devel
>= %{libtalloc_version}
46 Requires: libtalloc
>= %{libtalloc_version}
48 %if ! %with_included_tdb
49 BuildRequires
: libtdb
-devel
>= %{libtdb_version}
50 Requires: libtdb
>= %{libtdb_version}
52 %if ! %with_included_tevent
53 BuildRequires
: libtevent
-devel
>= %{libtevent_version}
54 Requires: libtevent
>= %{libtevent_version}
57 # To build the ctdb-pcp-pmda package, run rpmbuild with "--with pmda"
58 %define with_pcp_pmda
%{?_with_pmda: 1} %{?!_with_pmda: 0}
60 BuildRequires
: pcp
-libs
-devel
64 BuildRequires
: systemd
-units
68 ctdb is the clustered database used by samba
70 #######################################################################
76 # setup the init script and sysconfig file
77 %setup -T
-D
-n ctdb
-%{version} -q
82 if ccache
-h
>/dev
/null
2>&1 ; then
90 CFLAGS
="$RPM_OPT_FLAGS $EXTRA
-D_GNU_SOURCE" .
/buildtools
/bin
/waf configure \
91 --builtin-libraries
=replace
,popt \
92 --bundled
-libraries
=!talloc
,!tevent
,!tdb \
93 --minimum
-library
-version
=talloc
:%libtalloc_version,tdb
:%libtdb_version,tevent
:%libtevent_version \
98 --includedir
=%{_includedir}/ctdb \
100 --sysconfdir
=%{_sysconfdir} \
101 --mandir
=%{_mandir} \
102 --localstatedir
="
/var"
104 .
/buildtools
/bin
/waf build
107 # Clean up in case there is trash left from a previous build
108 rm -rf $RPM_BUILD_ROOT
110 # Create the target build directory hierarchy
111 mkdir -p $RPM_BUILD_ROOT
%{_sysconfdir}/sysconfig
112 mkdir -p $RPM_BUILD_ROOT
%{_sysconfdir}/sudoers.d
114 DESTDIR
=$RPM_BUILD_ROOT .
/buildtools
/bin
/waf
install
116 install -m644 config
/ctdb.sysconfig $RPM_BUILD_ROOT
%{_sysconfdir}/sysconfig
/ctdb
119 mkdir -p $RPM_BUILD_ROOT
%{_unitdir}
120 install -m
755 config
/ctdb.service $RPM_BUILD_ROOT
%{_unitdir}
122 mkdir -p $RPM_BUILD_ROOT
%{initdir}
123 install -m755 config
/ctdb.init $RPM_BUILD_ROOT
%{initdir}/ctdb
126 # This is a hack. All documents should be installed in /usr/share/doc.
127 rm -f $RPM_BUILD_ROOT
%{_sysconfdir}/ctdb
/events.d
/README
128 cp config
/events.d
/README README.eventscripts
129 rm -f $RPM_BUILD_ROOT
%{_sysconfdir}/ctdb
/nfs
-checks.d
/README
130 cp config
/nfs
-checks.d
/README README.nfs
-checks.d
131 cp config
/notify.d.README README.notify.d
133 # Remove "*.old" files
134 find $RPM_BUILD_ROOT
-name "
*.old"
-exec rm -f
{} \
;
137 rm -rf $RPM_BUILD_ROOT
140 #######################################################################
142 #######################################################################
145 %defattr(-,root
,root
)
147 %config(noreplace
) %{_sysconfdir}/sysconfig
/ctdb
148 %config(noreplace
) %{_sysconfdir}/ctdb
/notify.sh
149 %config(noreplace
) %{_sysconfdir}/ctdb
/debug
-hung
-script.sh
150 %config(noreplace
) %{_sysconfdir}/ctdb
/ctdb
-crash
-cleanup.sh
151 %config(noreplace
) %{_sysconfdir}/ctdb
/gcore_trace.sh
152 %config(noreplace
) %{_sysconfdir}/ctdb
/debug_locks.sh
155 %{_unitdir}/ctdb.service
157 %attr(755,root
,root
) %{initdir}/ctdb
160 %attr(755,root
,root
) %{_sysconfdir}/ctdb
/notify.d
162 %doc README COPYING NEWS
163 %doc README.eventscripts README.notify.d
164 %doc doc
/recovery
-process.txt
167 %{_sysconfdir}/sudoers.d
/ctdb
168 %{_sysconfdir}/ctdb
/functions
169 %{_sysconfdir}/ctdb
/events.d
/00.ctdb
170 %{_sysconfdir}/ctdb
/events.d
/01.reclock
171 %{_sysconfdir}/ctdb
/events.d
/05.system
172 %{_sysconfdir}/ctdb
/events.d
/10.interface
173 %{_sysconfdir}/ctdb
/events.d
/10.external
174 %{_sysconfdir}/ctdb
/events.d
/13.per_ip_routing
175 %{_sysconfdir}/ctdb
/events.d
/11.natgw
176 %{_sysconfdir}/ctdb
/events.d
/11.routing
177 %{_sysconfdir}/ctdb
/events.d
/20.multipathd
178 %{_sysconfdir}/ctdb
/events.d
/31.clamd
179 %{_sysconfdir}/ctdb
/events.d
/40.vsftpd
180 %{_sysconfdir}/ctdb
/events.d
/41.httpd
181 %{_sysconfdir}/ctdb
/events.d
/49.winbind
182 %{_sysconfdir}/ctdb
/events.d
/50.samba
183 %{_sysconfdir}/ctdb
/events.d
/60.nfs
184 %{_sysconfdir}/ctdb
/events.d
/70.iscsi
185 %{_sysconfdir}/ctdb
/events.d
/91.lvs
186 %{_sysconfdir}/ctdb
/events.d
/99.timeout
187 %config(noreplace
) %{_sysconfdir}/ctdb
/nfs
-checks.d
/00.portmapper.check
188 %config(noreplace
) %{_sysconfdir}/ctdb
/nfs
-checks.d
/10.status.check
189 %config(noreplace
) %{_sysconfdir}/ctdb
/nfs
-checks.d
/20.nfs.check
190 %config(noreplace
) %{_sysconfdir}/ctdb
/nfs
-checks.d
/30.nlockmgr.check
191 %config(noreplace
) %{_sysconfdir}/ctdb
/nfs
-checks.d
/40.mountd.check
192 %config(noreplace
) %{_sysconfdir}/ctdb
/nfs
-checks.d
/50.rquotad.check
193 %{_sysconfdir}/ctdb
/statd
-callout
194 %{_sysconfdir}/ctdb
/nfs
-linux
-kernel
-callout
196 %{_sbindir}/ctdbd_wrapper
198 %{_bindir}/ctdb_lock_helper
199 %{_bindir}/ctdb_event_helper
200 %{_bindir}/ctdb_recovery_helper
204 %{_bindir}/ctdb_diagnostics
208 %{_libdir}/libtevent
-unix
-util.so
.0*
209 %{_mandir}/man1
/ctdb
.1.gz
210 %{_mandir}/man1
/ctdbd
.1.gz
211 %{_mandir}/man1
/ctdbd_wrapper
.1.gz
212 %{_mandir}/man1
/onnode
.1.gz
213 %{_mandir}/man1
/ltdbtool
.1.gz
214 %{_mandir}/man1
/ping_pong
.1.gz
215 %{_mandir}/man5
/ctdbd.conf
.5.gz
216 %{_mandir}/man7
/ctdb
.7.gz
217 %{_mandir}/man7
/ctdb
-statistics
.7.gz
218 %{_mandir}/man7
/ctdb
-tunables
.7.gz
219 %{_libdir}/pkgconfig
/ctdb.pc
223 Summary: CTDB development libraries
224 Group: Development
/Libraries
227 development libraries
for ctdb
230 %defattr(-,root
,root
)
231 %{_includedir}/ctdb
/util
/*.h
232 %{_libdir}/libtevent
-unix
-util.so
235 Summary: CTDB
test suite
236 Group: Development
/Tools
237 Requires: ctdb
= %{version}
244 %defattr(-,root
,root
)
245 %dir %{_datadir}/%{name}-tests
246 %{_datadir}/%{name}-tests
/*
247 %dir %{_libdir}/%{name}-tests
248 %{_libdir}/%{name}-tests
/*
249 %{_bindir}/ctdb_run_tests
250 %{_bindir}/ctdb_run_cluster_tests
256 Summary: CTDB PCP pmda support
257 Group: Development
/Tools
258 Requires: ctdb
= %{version}
261 %description pcp
-pmda
262 Performance Co
-Pilot
(PCP
) support
for CTDB
265 %dir /var
/lib
/pcp
/pmdas
/ctdb
266 /var
/lib
/pcp
/pmdas
/ctdb
/*
273 * Thu Mar
1 2012 : Version 1.13
274 - This is the new stable branch
for modern features
for ctdb.
275 Main new features are performance
/scaling improvements
for
276 concurrnet fetch and fetch_lock operations.
277 * Tue Nov
8 2011 : Version 1.12
278 - Add new tunable
: AllowClientDBAttach that can be used to stop
279 client db access during maintenance operations
280 - Updated logging
for interfaces that are missing or dont exist but are
281 configured to be used.
282 - Add timeout argument to ctdb_cmdline_client
284 - Initial support
for '
readonly' delegations
for ctdb databases
285 This will when finished greatly improve performance
for contended hot
286 records that are used
for just
read-access.
287 - New 'ctdb cattdb'
command
288 - Massive updates to tests and eventscripts
289 - LCP2 ip allocation algorithm
290 - Record Fetch collapse. Collapse multiple fetch
-lock requests from cients
291 to a single network fetch and defer other concurrent requests
until the
292 initial fetch completes
, and
then service the deferred calls locally.
293 This will greatly improve performance
for contended hot records
294 where clients request write
-locks.
295 * Thu Sep
1 2011 : Version 1.11
297 * Tue May
25 2010 : Version 1.10
299 * Tue May
25 2010 : Version 1.9
301 * Thu Mar
25 2010 : Version 1.0.114
302 - Lots of changes from Metze
303 * Wed Jan
13 2010 : Version 1.0.113
304 - Incorrect use of dup2
() could cause ctdb to spin eating
100% cpu.
305 * Tue Jan
12 2010 : Version 1.0.112
306 - Revert the use of wbinfo
--ping
-dc as it is proving too unreliable.
307 - Minor testsuite changes.
308 * Fri Dec
18 2009 : Version 1.0.111
309 - Fix a logging bug when an eventscript is aborted that could cause a crash.
310 - Add back cb_status that was lost
in a previous commit.
311 * Fri Dec
18 2009 : Version 1.0.110
312 - Metxe
: fix
for filedescriptor leak
in the new eventscript code.
313 - Rusty
: fix
for a crash bug
in the eventscript code.
314 * Thu Dec
17 2009 : Version 1.0.109
315 - Massive eventscript updates.
(bz58828
)
316 - Nice the daemon instead of using realtime scheduler
, also use mlockall
() to
317 reduce the risk of blockign due to paging.
318 - Workarounds
for valgrind when forking once
for each script. Valgrind consumes
319 massive cpu when terminating the scripts on virtual systems.
320 - Sync the tdb library with upstream
, and use the new TDB_DISALLOW_NESTING flag.
321 - Add new
command "ctdb dumpdbbackup"
322 - Start using the new tdb check framework to validate tdb files upon startup.
323 - A new framework where we can control health
for individual tdb databases.
324 - Fix a crash bug
in the logging code.
325 - New transaction code
for persistent databases.
326 - Various other smaller fixes.
327 * Mon Dec
7 2009 : Version 1.0.108
328 - Transaction updates from Michael Adam.
329 - Use the new wbinfo
--ping
-dc instead of
-p
in the eventscript
for samba
330 the check
if winbindd is ok.
331 - Add a better "process
-exist"
for samba so it will automatically
332 reap smbd's on stopped and banned nodes to reclaim subrecords.
333 This will be
done a bit differently
in the next release.
334 - Use a statically allocated buffer
for the 'first
-time' capture buffer
335 to reduce the pressure on malloc
/free.
336 * Wed Dec
2 2009 : Version 1.0.107
337 - fix
for rusty to solve a double
-free that can happen when there are
338 multiple packets queued and the connection is destroyed before
339 all packets are processed.
340 * Tue Dec
1 2009 : Version 1.0.106
341 - Buildscript changes from Michael Adam
342 - Dont
do a full recovery when there is a mismatch detected
for ip addresses
,
343 just
do a less disruptive ip
-reallocation
344 - When starting ctdbd
, wait until all initial recoveries have finished
345 before we issue the "startup" event.
346 So dont start services or monitoring
until the cluster has
348 - Major eventscript overhaul by Ronnie
, Rusty and Martins and fixes of a few
350 * Thu Nov
19 2009 : Version 1.0.105
351 - Fix a bug where we could SEGV
if multiple concurrent "ctdb eventscript ..."
352 are used and some of them block.
353 - Monitor the daemon from the syslog child process so we shutdown cleanly when
354 the main daemon terminates.
355 - Add a
500k line ringbuffer
in memory where all log messages are stored.
356 - Add a "ctdb getlog
<level
>"
command to pull log messages from the
in memory
358 - From martin
: fixes to cifs and nfs autotests
359 - from michael a
: fix a bashism
in 11.natgw
360 * Fri Nov
6 2009 : Version 1.0.104
361 - Suggestion from Metze
, we can now use killtcp to
kill local connections
362 for nfs so change the killtcp script to
kill both directions of an NFS
364 We used to deliberately only
kill one direction
in these cases due to
366 - Suggestion from christian Ambach
, when using natgw
, try to avoid using a
367 UNHEALTHY node as the natgw master.
368 - From Michael Adam
: Fix a SEGV bug
in the recent change to the eventscripts
369 to allow the timeout to apply to each individual script.
370 - fix a talloc bug
in teh vacuuming code that produced nasty valgrind
372 - From Rusty
: Set up
ulimit to create core files
for ctdb
, and spawned
373 processes by default. This is useful
for debugging and testing but can be
374 disabled by setting CTDB_SUPRESS_COREFILE
=yes
in the sysconfig file.
375 - Remove the wbinfo
-t check from the startup check that winbindd is happy.
376 - Enhance the
test for bond devices so we also check
if the sysadmin have
377 disabled all slave devices using "ifdown".
378 * Tue Nov
3 2009 : Version 1.0.103
379 - Dont use vacuuming on persistent databases
380 - Michael A
: transaction updates to persistent databases
381 - Dont activate service automatically when installing the RPM. Leave this to the admin.
382 - Create a child process to send all log messages to
, to prevent a hung
/slow syslogd
383 from blocking the main daemon. In this
case, discard log messages instead and
let the child
385 - Michael A
: updates to log messages
386 * Thu Oct
29 2009 : Version 1.0.102
387 - Wolfgang
: fix
for the vacuuming code
388 - Wolfgang
: stronger tests
for persistent database filename tests
389 - Improve the log message when we refuse to startup since wbinfo
-t fails
390 to
make it easier to spot
in the log.
391 - Update the uptime
command output and the man page to indicate that
392 "
time since last ..."
if from either the last recovery OR the last failover
393 - Michael A
: transaction updates
394 * Wed Oct
28 2009 : Version 1.0.101
395 - create a separate context
for non
-monitoring events so they dont interfere with the monitor event
396 - make sure to
return status
0 in teh callback when we abort an event
397 * Wed Oct
28 2009 : Version 1.0.100
398 - Change eventscript handling to allow EventScriptTimeout
for each individual script instead of
for all scripts as a whole.
399 - Enhanced logging from the eventscripts
, log the name and the duration
for each script as it finishes.
400 - Add a check to use wbinfo
-t
for the startup event of samba
401 - TEMP
: allow clients to attach to databases even when teh node is
in recovery mode
402 - dont run the monitor event as frequently after an event has failed
403 - DEBUG
: in the eventloops
, check the
local time and warn
if the
time changes backward or rapidly forward
404 - From Metze
, fix a bug where recovery master becoming unhealthy did not trigger an ip failover.
405 - Disable the multipath script by default
406 - Automatically re
-activate the reclock checking
if the reclock file is specified at runtime. Update manpage to reflect this.
407 - Add a mechanism where samba can register a SRVID and
if samba unexpectedly disconnects
, a message will be broadcasted to all other samba daemons.
408 - Log the pstree on hung scripts to a file
in /tmp isntead of
/var
/log
/messages
409 - change ban count before unhealthy
/banned to
10
410 * Thu Oct
22 2009 : Version 1.0.99
411 - Fix a SEGV
in the new db priority code.
412 - From Wolfgang
: eliminate a ctdb_fatal
() if there is a dmaster violation detected.
413 - 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.
414 - Lower the debuglevel
for the messages when ctdb creates a filedescruiptor so we dont spam the logs with these messages.
415 - Dont have the RPM automatically restart ctdb
416 - Volker
: add a missing transaction_cancel
() in the handling of persistent databases
417 - 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.
418 * Tue Oct
20 2009 : Version 1.0.98
419 - Fix
for the vacuuming database from Wolfgang M
420 - Create a directory where the
test framework can put temporary overrides
421 to variables and functions.
422 - Wait a lot longer before shutting down the node when the reclock file
423 is incorrectly configured
, and log where it is configured.
424 - Try to avoid running the "monitor" event when databases are frozen.
425 - Add logging
for every
time we create a filedescriptor so we can
trap
427 * Wed Oct
14 2009 : Version 1.0.97
428 - From martins
: update onnode.
429 Update onnode to allow specifying an alternative nodes file from
430 the
command line and also to be able to specify hostnames on the
432 onnode host1
,host2
,...
433 * Tue Oct
13 2009 : Version 1.0.96
434 - Add more debugging output when eventscripts have trouble. Print a
435 "pstree
-p" to the log when scripts have hung.
436 - Update the initscript
, only print the "No reclock file used" warning
437 when we
do "service ctdb start"
, dont also print them
for all other
439 - When changing between unhealthy
/healthy state
, push a request to the
440 recovery master to perform an ip reallocation instead of waiting
for the
441 recovery master to pull and check the state change.
442 - Fix a bug
in the new db
-priority handling where a pre
-.95 recovery master
443 could no longer lock the databases on a post
-.95 daemon.
444 - Always create the nfs state directories during the "monitor" event.
445 This makes it easier to configure and
enable nfs at runtime.
446 - From Volker
, forward
-port a simper deadlock avoiding patch from the
1.0.82
447 branch. This is a simpler versionof the "db priority lock order" patch
448 that went into
1.0.95, and will be kept
for a few versions
until samba
449 has been updated to use the functionality from
1.0.95.
450 * Mon Oct
12 2009 : Version 1.0.95
451 - Add database priorities. Allow samba to
set the priority of databases
452 and lock the databases
in priority order during recovery
453 to avoid a deadlock when samba locks one database
then blocks indefinitely
454 while waiting
for the second databaso to become locked.
455 - Be aggressive and ban nodes where the recovery transaction start call
457 * Sat Oct
10 2009 : Version 1.0.94
458 - Be very aggressive and quickly ban nodes that can not freeze their databases
459 * Thu Oct
8 2009 : Version 1.0.93
460 - When adding an ip
, make sure to update this assignment on all nodes
461 so it wont show up as
-1 on other nodes.
462 - When adding an ip and immediately deleting it
, it was possible that
463 the daemon would crash accessing already freed memory.
464 Readjust the memory hierarchy so the destructors are called
in the right order.
465 - Add a handshake to the recovery daemon to eliminate some rare cases where
466 addip
/delip might cause a recovery to occur.
467 - updated onnode documenation from Martin S
468 - Updates to the natgw eventscript to allow disabling natgw at runtime
469 * Fri Oct
2 2009 : Version 1.0.92
470 - Test updates and merge from martin
471 - Add notification
for "startup"
472 - Add documentation
for notification
473 - from martin
, a fix
for restarting vsftpd
in the eventscript
474 * Tue Sep
29 2009 : Version 1.0.91
475 - New vacuum and repack design from Wolgang Mueller.
476 - 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.
477 - Add machinereadable output to the ctdb getreclock
command
478 - merge transaction updates from Michael Adam
479 - In the new banning code
, reset the culprit count to
0 for all nodes that could successfully compelte a full recovery.
480 - 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.
481 * Sat Sep
12 2009 : Version 1.0.90
482 - Be more forgiving
for eventscripts that hang during startup
483 - Fix
for a banning bug
in the new banning logic
484 * Thu Sep
3 2009 : Version 1.0.89
485 - Make it possible to manage winbind independently of samba.
486 - Add new prototype banning code
487 - Overwrite the vsftpd state file instead of appending. This eliminates
488 annoying errors
in the log.
489 - Redirect some iptables commands to dev null
490 - From Michael A
, explicitely
set the broadcast when we takeover a public ip
491 - Remove a reclock file check we no longer need
492 - Skip any persistent database files ending
in .bak
493 * Mon Aug
17 2009 : Version 1.0.88
494 - Add a new state
for eventscripts
: DISABLED.
495 Add two new commands "ctdb enablescript
/disablescript" to
enable/disable
496 eventscripts at runtime.
497 - Bugfixes
for TDB from rusty.
498 - Merge
/Port changes from upstream TDB library by rusty.
499 - Additional new tests from MartinS. Tests
for stop
/continue.
500 - Initial patch to rework vacuuming
/repacking process from Wolfgang Mueller.
501 - Updates from Michael Adam
for persistent writes.
502 - Updates from MartinS to handle the new STOPPED bit
in the
test framework.
503 - Make it possible to
enable/disable the RECMASTER and LMASTER roles
504 at runtime. Add two new commands
505 "ctdb setlmasterrole
/setrecmasterrole on
/off"
506 - Make it possible to
enable/disable the natgw feature at runtime. Add
507 the
command "ctdb setnatgwstate on
/off"
508 * Fri Jul
17 2009 : Version 1.0.87
509 - Add a new event "stopped" that is called when a node is stopped.
510 - Documentation of the STOPPED flag and the stop
/continue commands
511 - Make it possible to start a node
in STOPPED mode.
512 - Add a new node flag
: STOPPED and commands "ctdb stop" "ctdb
continue"
513 These commands are similar to "diasble
/enable" but will also remove the node from the vnnmap
, while disable only fails all ip addresses over.
514 - tests
for NFS
, CIFS by martins
515 - major updates to the init script by martins
516 - Send gratious arps with a
1.1 second stride instead of a
1 second stride to workaround interesting "features" of common linux stacks.
517 - Various
test enhancements from martins
:
518 - additional other tests
519 - add tests
for grat arp generation
, ping during failover
, ssh and failover
520 - New
/updated tcp tickle tests and supprot functions
521 - provide better debugging when a
test fails
522 - make ctdbd restarts more reliable
in the tests
523 - update the "
wait bar" to
make the
wait progress
in tests more obvious
525 - 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.
526 - document the ipreallocate
command
527 - Updates to
enable/disable to use the ipreallocate
command to block
until the following ipreallocation has completed.
528 - Update the main daemon and the tools to allow debug level to be a string instead of an integer.
529 - Update the sysconfig file to show using string literals instead of numeric values
for the debuglevels used.
530 - If no debuglevel is specific
, make "ctdb setdebug" show the available options.
531 - 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.
532 * Tue Jun
30 2009 : Version 1.0.86
533 - Do not access the reclock at all
if VerifyRecoveryLock is zero
, not even try to probe it.
534 - Allow setting the reclock file as ""
, which means that no reclock file at all should be used.
535 - Document that a reclock file is no longer required
, but that it is dangerous.
536 - Add a control that can be used to
set/clear
/change the reclock file
in the daemon during runtime.
537 - Update the recovery daemon to poll whether a reclock file should be sued and
if so which file at runtime
in each monitoring cycle.
538 - Automatically disable VerifyRecoveryLock everytime a user changes the location of the reclock file.
539 - do not allow the VerifyRecoveryLock to be
set using ctdb setvar
if there is no recovery lock file specified.
540 - Add two commands "ctdb getreclock" and "ctdb setreclock" to modify the reclock file.
541 * Tue Jun
23 2009 : Version 1.0.85
542 - From William Jojo
: Dont use getopt on AIX
543 - Make it possible to use "ctdb listnodes" also when the daemon is not running
544 - Provide machinereadable output to "ctdb listnodes"
545 - Dont list DELETED nodes
in the ctdb listnodes output
546 - Try to avoid causing a recovery
for the average
case when adding
/deleting
/moving an ip
547 - When banning a node
, drop the IPs on that node only and not all nodes.
548 - Add tests
for NFS and CIFS tickles
549 - Rename
99.routing to
11.routing so it executes before NFS and LVS scripts
550 - Increase the default timeout before we deem an unresponsive recovery daemon hung and shutdown
551 - Reduce the reclock timout to
5 seconds
552 - Spawn a child process
in the recovery daemon ot check the reclock file to
553 avoid blocking the process
if the underlying filesystem is unresponsive
554 - fix
for filedescriptor leak when a child process timesout
555 - Dont log errors
if waitpid
() returns
-1
556 - Onnode updates by Martins
557 - Test and initscript cleanups from Martin S
558 * Tue Jun
2 2009 : Version 1.0.84
559 - Fix a bug
in onnode that could not handle dead nodes
560 * Tue Jun
2 2009 : Version 1.0.83
561 - Document how to remove a ndoe from a running cluster.
562 - Hide all deleted nodes from ctdb output.
563 - Lower the loglevel on some eventscript related items
564 - Dont queue packets to deleted nodes
565 - When building initial vnnmap
, ignode any nonexisting nodes
566 - Add a new nodestate
: DELETED that is used when deleting a node from an
568 - dont remove the ctdb socket when shutting down. This prevents a race
in the
569 initscripts when restarting ctdb quickly after stopping it.
570 - TDB nesting reworked.
571 - Remove obsolete ipmux
572 - From Flavio Carmo Junior
: Add eventscript and documentation
for ClamAV antivirus engine
573 - From Sumit Bose
: fix the regex
in the
test to handle the new ctdb
574 statistics output that was recently added.
575 - change the socket
type we use
for grauitious arps from the obsolete
576 AF_INET
/SOCK_PACKET to instead use PF_PACKET
/SOCK_RAW.
577 - Check
return codes
for some functions
, from Sumit Bose
, based on codereview by Jim Meyering.
578 - Sumit Bose
: Remove structure memeber node_list_file that is no longer used.
579 - Sumit Bose
: fix configure warning
for netfilter.h
580 - Updates to the webpages by Volker.
581 - Remove error messages about missing
/var
/log
/log.ctdb file from ctdb_diagnostics.sh from christian Ambach
582 - Additional error logs
if hte eventscript switching from dameon to client mode fails.
583 - 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.
584 * Thu May
14 2009 : Version 1.0.82
585 - Update the "ctdb lvsmaster"
command to
return -1 on error.
586 - Add a
-Y flag to "ctdb lvsmaster"
587 - RHEL5 apache leaks semaphores when crashing. Add semaphore cleanup to the
588 41.httpd eventscript and try to restart apache when it has crashed.
589 - Fixes to some tests
590 - Add a
-o option to "onnode" which will redirect all stdout to a file
for
592 - Add a natgw and a lvs node specifier to onnode so that we can use
594 - Assign the natgw address to lo instead of the private network so it can also
595 be used where private and public networks are the same.
596 - Add GPL boilerplates to two missing scripts.
597 - Change the natgw prefix NATGW_ to CTDB_NATGW_
598 * Fri May
8 2009 : Version 1.0.81
599 - use smbstatus
-np instead of smbstatus
-n
in the
50.samba eventscript
600 since this avoids performing an expensive traverse on the locking and brlock
602 - make ctdb automatically terminate all traverse child processes clusterwide
603 associated to a client application that terminates before the traversal is
605 - From Sumit Bose
: fixes to AC_INIT handling.
606 - From Michael Adam
, add Tridge's "ping_pong" tool the the ctdb distro since
607 this is very useful
for testing the backend filesystem.
608 - From Sumit bose
, add support
for additional
64 bit platforms.
609 - Add a link from the webpage to Michael Adams SambaXP paper on CTDB.
610 * Fri May
1 2009 : Version 1.0.80
611 - change init shutdown level to
01 for ctdb so it stops before any of the other services
612 - if we can not pull a database from a remote node during recovery
, mark that node as a culprit so it becomes banned
613 - 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
614 - 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.
615 - dont unconditionally
kill off ctdb and restrat it on "service ctdb start". Fail "service ctdb start" with an error
if ctdb is already running.
616 - 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.
617 - fix a cosmetic bug with ctdb statistics where certain counters could become negative.
618 * Wed Apr
8 2009 : Version 1.0.79
619 - From Mathieu Parent
: add a ctdb pkgconfig file
621 - add a funciton remove_ip to safely remove an ip from an interface
, taking care to workaround an issue with linux
alias interfaces.
622 - Update the natgw eventscript to use the safe remove_ip
() function
623 - fix a bug
in the eventscript child process that would cause the socket to be removed.
624 - dont verify nodemap on banned nodes during cluster monitoring
625 - Update the dodgy SeqnumInterval to have ms resolution
626 * Tue Mar
31 2009 : Version 1.0.78
627 - Add a notify mechanism so we can send snmptraps
/email to external management systems when the node becomes unhealthy
628 - include
11.natgw eventscript
in thew
install so that the NATGW feature works
629 * Tue Mar
31 2009 : Version 1.0.77
630 - 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
631 * Wed Mar
25 2009 : Version 1.0.76
632 - Add a debugging
command "xpnn" which can print the pnn of the node even when ctdbd is not running.
633 - Redo the NATGW implementation to allow multiple disjoing NATGW groups
in the same cluster.
634 * Tue Mar
24 2009 : Version 1.0.75
635 - Various updates to LVS
636 - Fix a bug
in the killtcp control where we did not
set the port correctly
637 - add a new "ctdb scriptstatus"
command that shows the status of the eventrscripts.
638 * Mon Mar
16 2009 : Version 1.0.74
639 - Fixes to AIX from C Cowan.
640 - Fixes to ctdb_diagnostics so we collect correct GPFS data
641 - Fixes to the net conf list
command in ctdb_diagnostics
642 - Check the static
-routes file IFF it exists
in ctdb_diagnostics
643 * Wed Mar
4 2009 : Version 1.0.73
644 - Add possibility to disable the check of shares
for NFS and Samba
645 - From Sumit Bose
, fix dependencies so
make -j works
646 * Wed Feb
18 2009 : Version 1.0.72
647 - Updates to
test scripts by martin s
648 - Adding a COPYING file
649 - Use netstat to check
for services and ports and fallback to netcat
650 only
if netstat is unavailable.
651 * Mon Feb
2 2009 : Version 1.0.71
652 - Additional ipv6 fixes from Michael Adams
653 * Fri Jan
16 2009 : Version 1.0.70
654 - IPv6 support is completed. this is backward compatible with ipv4
-only
655 systems. To use IPv6 with samba and ctdb you need current GIT of samba
3.3
656 or michael adams samba
-ctdeb branch.
657 - Many enhancements to the build system and scripts to
make it more SUSE
658 friendly by Michael Adams.
659 - Change of how the naming of the package is structured. We are now
660 using "
1.0.70" as a release and "
-1" as the revision instead of as
661 previously using "
1.0" as release and "
.70" as the revision.
663 * Thu Dec
18 2008 : Version 1.0.69
664 - Various fixes to scripts by M Adam
665 - Dont call ctdb_fatal
() when the transport is down during shutdown
666 * Fri Dec
12 2008 : Version 1.0.68
667 - Fixes
for monitoring of interfaces status from Michael Adam.
668 - Use
-q instead of
>/dev
/null
for grep to enhance readability of the
669 scripts from Michael Adam.
670 - Update to the "ctdb recover"
command. This
command now block
until the
671 has completed. This makes it much easier to use
in scripts and avoids
672 the common workaround
:
674 ... loop
while waiting
for recovery completes ...
676 - Add a CTDB_TIMEOUT variable. If
set, this variable provides an automatic
677 timeout
for "ctdb
<command>"
, similar to using
-T
<timeout
>
678 - Set a unique errorcode
for "ctdb
<command>" when it terminates due to a
679 timeout so that scripts can distinguish between a hung
command and what was
681 - Update "ctdb ban
/unban" so that
if the cluster is
in recovery these commands
682 blocks and waits
until after recovery is complete before the perform the
683 ban
/unban operation. This is necessary since the recovery process can cause
684 nodes to become automatically unbanned.
685 - Update "ctdb ban
/unban" to block
until the recovery that will follow shortly
686 after this
command has completed.
687 This makes it much easier to use
in scripts and avoids the common
690 ... loop
while waiting
for recovery completes ...
692 - Bugfix
for the new flags handling
in 1.0.67. Abort and restart monitoring
693 if we failed to get proper nodemaps from a remote node instead of
694 dereferencing a null pointer.
695 - If ctdbd was explicitely started with the '
--socket' argument
, make
696 ctdbd automatically
set CTDB_SOCKET to the specified argument.
697 This ensures that eventscripts spawned by the ctdb daemon will default to
698 using the same socket and talk to the correct daemon.
699 This primarily affects running multiple daemons on the same host and where
700 you want each instance of ctdb daemons have their eventscripts talk to the
702 - Update "ctdb ping" to
return an error code
if the ping fail so that it
703 can be used
in scripts.
704 - Update to how to synchronize management of node flags across the cluster.
705 * Wed Dec
3 2008 : Version 1.0.67
706 - Add a document describing the recovery process.
707 - Fix a bug
in "ctdb setdebug" where it would refuse to
set a negative
709 - Print the list of literals
for debug names
if an invalid one was given
711 - Redesign how "ctdb reloadnodes" works and reduce the amont of tcp teardowns
712 used during this event.
713 - Make it possible to delete a public ip from all nodes at once using
715 * Mon Nov
24 2008 : Version 1.0.66
716 - Allow to change the recmaster even when we are not frozen.
717 - Remove two redundant SAMBA_CHECK variables from the sysconf example
718 - After a node failure it can take very long before some lock operations
719 ctdb needs to perform are allowed
/works with gpfs again. Workaround this
720 by treating a hang
/timeout as success.
721 - Dont override CTDB_BASE is fet
in the shell already
722 - Always send keepalive packets regardless of whether the link is idle or not.
723 - Rewrite the disable
/enable flag update logic to prevent a race between
724 "ctdb disable
/enable" and the recovery daemon when updating the flags to
726 * Thu Nov
13 2008 : Version 1.0.65
727 - Update the sysconfig example
: The default debug level is
2 (NOTICE
) and not
729 - Add support
for a CTDB_SOCKET environment variable
for the ctdb
command
730 line tool. If
set, this overrides the default socket the ctdb tool will
732 - Add logging of high latency operations.
733 * Wed Oct
22 2008 : Version 1.0.64
734 - Add a context and a timed event so that once we have been
in recovery
for
735 too long we drop all public addresses.
736 * Mon Oct
20 2008 : Version 1.0.63
737 - Remove logging of "periodic cleanup ..."
in 50.samba
738 - When we reload a nodes file
, we must detect this and reload the file also
739 in the recovery daemon before we try to dereference somethoung beyond the end
741 * Thu Oct
16 2008 : Version 1.0.62
742 - Allow multiple eventscritps using the same prefix number.
743 It is undefined which order scripts with the same prefix will execute
in.
744 * Wed Oct
15 2008 : Version 1.0.61
745 - Use "route add
-net" instead of "ip route add" when adding routes
in 99.routing
746 - lower the loglevel os several debug statements
747 - check the status returned from ctdb_ctrl_get_tickles
() before we try to print them out to the screen.
748 - install a new eventscript
20.multipathd whoich can be used to monitor that multipath devices are healthy
749 * Wed Oct
15 2008 : Version 1.0.60
750 - Verify that nodes we try to ban
/unban are reachable and print an error othervise.
751 - 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.
752 - From Mathieu Parent
: log debian systems log the package versions
in ctdb diagnostics
753 - From Mathieu Parent
: specify logdir location
for debian
(this patch was later reversed
)
754 - From Michael Adams
: allow
# comments in nodes/public_addresses files
755 * Tue Oct
7 2008 : Version 1.0.59
756 - Updated "reloadnodes" logic. Instead of bouncing the entire tcp layer it is sufficient to just close and reopen all outgoing tcp connections.
757 - 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
)
758 - IDR tree fix from Jim Houston
759 - Better handling of critical events
if the
local clock is suddenly changed forward by a lot.
760 - Fix three slow memory leaks
in the recovery daemon
761 - New ctdb
command : ctdb recmaster which prints the pnn of the recmaster
762 - Onnode enhancements from Martin S
: "healthy" and "connected" are now possible nodespecifiers
763 - From Martin S
: doc fixes
764 - lowering some debug levels
for some nonvital informational messages
765 - Make the daemon daemon monitoring stronger and allow ctdbd to detect a hung
767 - From C Cowan
: patches to compile ipv6 under AIX
768 - zero out some structs to keep valgrind happy
769 * Wed Aug
27 2008 : Version 1.0.58
770 - revert the name change tcp_tcp_client back to tcp_control_tcp so
772 - Updates to the init script from Abhijith Das
<adas@redhat.com
>
773 * Mon Aug
25 2008 : Version 1.0.57
774 - initial support
for IPv6
775 * Mon Aug
11 2008 : Version 1.0.56
776 - fix a memory leak
in the recovery daemon.
777 * Mon Aug
11 2008 : Version 1.0.55
778 - Fix the releaseip message we seond to samba.
779 * Fri Aug
8 2008 : Version 1.0.54
780 - fix a looping error
in the transaction code
781 - provide a more detailed error code
for persistent store errors
782 so clients can
make more intelligent choices on how to try to recover
783 * Thu Aug
7 2008 : Version 1.0.53
784 - Remove the reclock.pnn file it can cause gpfs to fail to umount
785 - New transaction code
786 * Mon Aug
4 2008 : Version 1.0.52
787 - Send an explicit gratious arp when starting sending the tcp tickles.
788 - When doing failover
, issue a killtcp to non
-NFS
/non
-CIFS clients
789 so that they fail quickly. NFS and CIFS already fail and recover
791 - Update the
test scripts to handle CTRL
-C to
kill off the
test.
792 * Mon Jul
28 2008 : Version 1.0.51
793 - Strip off the vlan tag from bond devices before we check
in /proc
794 if the interface is up or not.
795 - Use testparm
in the background
in the scripts to allow probing
796 that the shares
do exist.
797 - Fix a bug
in the logging code to handle multiline entries better
798 - Rename private elements from private to private_data
799 * Fri Jul
18 2008 : Version 1.0.50
800 - Dont assume that just because we can establish a TCP connection
801 that we are actually talking to a functioning ctdb daemon.
802 So dont mark the node as CONNECTED just because the tcp handshake
804 - Dont try to
set the recmaster to ourself during elections
for those
805 cases we know this will fail. To remove some annoying benign but scary
806 looking entries from the log.
807 - Bugfix
for eventsystem
for signal handling that could cause a node to
809 * Thu Jul
17 2008 : Version 1.0.49
810 - Update the safe persistent update fix to work with unpatched samba
812 * Thu Jul
17 2008 : Version 1.0.48
813 - Update the spec file.
814 - Do not start new user
-triggered eventscripts
if we are already
815 inside recovery mode.
816 - Add two new controls to start
/cancel a persistent update.
817 A client such as samba can use these to tell ctdbd that it will soon
818 be writing directly to the persistent database tdb file. So
if
819 samba is
-9ed before it has eitehr
done the persistent_store or
820 canceled the operation
, ctdb knows that the persistent databases
821 'may' be out of sync and therefore a full blown recovery is called
for.
822 - Add two new options
:
823 CTDB_SAMBA_SKIP_CONF_CHECK and CTDB_SAMBA_CHECK_PORTS that can be used
824 to override what checks to
do when monitoring samba health.
825 We can no longer use the smbstatus
, net or testparm commands to check
826 if samba or its config is healthy since these commands may block
827 indefinitely and thus can not be used
in scripts.
828 * Fri Jul
11 2008 : Version 1.0.47
829 - Fix a double free bug where
if a user striggered
(ctdb eventscript
)
830 hung and
while the timeout handler was being processed a new user
831 triggered eventscript was started we would free state twice.
832 - Rewrite of onnode and associated documentation.
833 * Thu Jul
10 2008 : Version 1.0.46
834 - Document both the LVS
:cingle
-ip
-address and the REMOTE
-NODE
:wan
-accelerator
836 - Add commands "ctdb pnn"
, "ctdb lvs"
, "ctdb lvsmaster".
837 - LVS improvements. LVS is the single
-ip
-address mode
for a ctdb cluster.
838 - Fixes to supress rpmlint warnings
840 - Change \s to
[[:space
:]] in some scripts. Not all RHEL5 packages come
841 with a egrep that handles \s even same version but different arch.
842 - Revert the change to NFS restart. CTDB should NOT attempt to restart
844 - Rewrite of the waitpid
() patch to use the eventsystem
for handling
846 * Tue Jul
8 2008 : Version 1.0.45
847 - Try to restart the nfs service
if it has failed to respond
3 times in a row.
848 - waitpid
() can block
if the child does not respond promptly to SIGTERM.
849 ignore all SIGCHILD signals by setting SIGCHLD to SIG_DEF.
850 get rid of all calls to waitpid
().
851 - make handling of eventscripts hanging more liberal.
852 only consider the script to have failed and making the node unhealthy
853 IF the eventscript terminated wiht an error
854 OR the eventscript hung
5 or more
times in a row
855 * Mon Jul
7 2008 : Version 1.0.44
856 - Add a CTDB_VALGRIND option to
/etc
/sysconfig
/ctdb to
make it start
857 ctdb under valgrind. Logs go to
/var
/log
/ctdb_valgrind.PID
858 - Add a hack to show the control opcode that caused uninitialized data
859 in the valgrind output by encoding the opcode as the line number.
860 - Initialize structures and allocated memory
in various places
in
861 ctdb to
make it valgrind
-clean and remove all valgrind errors
/warnings.
862 - If
/when we destroy a lockwait child
, also
make sure we cancel any pending transactions
863 - If a transaction_commit fails
, delete
/cancel any pending transactions and
864 return an error instead of calling ctdb_fatal
()
865 - When running ctdb under valgrind
, make sure we run it with
--nosetsched and also
866 ensure that we
do not use mem
-mapped i
/o when accessing the tdb's.
867 - zero out ctdb
->freeze_handle when we free
/destroy a freeze
-child.
868 This prevent a heap corruption
/ctdb crash bug that could trigger
869 if the freeze child
times out.
870 - we dont need to explicitely thaw the databases from the recovery daemon
871 since this is
done implicitely when we restore the recovery mode back to normal.
872 - track when we start and stop a recovery. Add the '
time it took to complete the
873 recovery' to the 'ctdb uptime' output.
874 Ensure by tracking the start
/stop recovery timestamps that we
do not
875 check that the ip allocation is consistend from inside the recovery daemon
876 while a different node
(recovery master
) is performing a recovery.
877 This prevent a race that could cause a full recovery to trigger
if the
878 'ctdb disable
/enable' commands took very long.
879 - The freeze child indicates to the master daemon that all databases are locked
880 by writing data to the pipe shared with the master daemon.
881 This write sometimes fail and thus the master daemon never notices that the databases
882 are locked cvausing long timeouts and extra recoveries.
883 Check that the write is successful and try the write again
if it failed.
884 - In each node
, verify that the recmaster have the right node flags
for us
885 and force a push of our flags to the recmaster
if wrong.
886 * Tue Jul
1 2008 : Version 1.0.43
887 - Updates and bugfixes to the specfile to keep rpmlint happy
888 - Force a global flags update after each recovery event.
889 - Verify that the recmaster agrees with our node flags and update the
891 - When writing back to the parent from a freeze
-child across the pipe
,
892 loop over the write
in case the write failed with an error othervise
893 the parent will never be notified tha the child has completed the operation.
894 - Automatically thaw all databases when recmaster marks us as being
in normal
895 mode instead of recovery mode.
896 * Fri Jun
13 2008 : Version 1.0.42
897 - When event scripts have hung
/timedout more than EventScriptBanCount
times
898 in a row the node will ban itself.
899 - Many updates to persistent write tests and the
test scripts.
900 * Wed May
28 2008 : Version 1.0.41
901 - Reactivate the safe writes to persistent databases and solve the
902 locking issues. Locking issues are solved the only possible way
,
903 by using a child process to
do the writes. Expensive and slow but... .
904 * Tue May
27 2008 : Version 1.0.40
905 - Read the samba sysconfig file from the
50.samba eventscript
906 - Fix some emmory hierarchical bugs
in the persistent write handling
907 * Thu May
22 2008 : Version 1.0.39
908 - Moved a CTDB_MANAGES_NFS
, CTDB_MANAGES_ISCSI and CTDB_MANAGES_CSFTPD
909 into
/etc
/sysconfig
/ctdb
910 - Lowered some debug messages to not fill the logfile with entries
911 that normally occur
in the default configuration.
912 * Fri May
16 2008 : Version 1.0.38
913 - Add machine readable output support to "ctdb getmonmode"
914 - Lots of tweaks and enhancements
if the event scripts are "slow"
915 - Merge from tridge
: an attempt to
break the chicken
-and
-egg deadlock that
916 net conf introduces
if used from an eventscript.
917 - Enhance tickles so we can tickle an ipv6 connection.
918 - Start adding ipv6 support
: create a new container to replace sockaddr_in.
919 - Add a checksum routine
for ipv6
/tcp
920 - When starting up ctdb
, let the init script
do a tdbdump on all
921 persistent databases and verify that they are good
(i.e. not corrupted
).
922 - Try to use "safe transactions" when writing to a persistent database
923 that was opened with the TDB_NOSYNC flag. If we can get the transaction
924 thats great
, if we cant we have to write anyway since we cant block here.
925 * Mon May
12 2008 : Version 1.0.37
926 - When we shutdown ctdb we close the transport down before we run the
927 "shutdown" eventscripts. If ctdb decides to send a packet to a remote node
928 after we have shutdown the transport but before we have shutdown ctdbd
929 itself this could lead to a SEGV instead of a clean shutdown. Fix.
930 - When using the "exportfs"
command to extract which NFS
export directories
931 to monitor
, exportfs violates the "principle of least surprise" and
932 sometimes report a single
export line as two lines of text output
933 causing the monitoring to fail.
934 * Fri May
9 2008 : Version 1.0.36
935 - fix a memory corruption bug that could cause the recovery daemon to crash.
936 - fix a bug with distributing public ip addresses during recovery.
937 If the node that is the recovery master did NOT use public addresses
,
938 then it assumed that no other node
in the cluster used them either and
939 thus skipped the entire step of reallocating public addresses.
940 * Wed May
7 2008 : Version 1.0.35
941 - During recovery
, when we define the new
set of lmasters
(vnnmap
)
942 only consider those nodes that have the can
-be
-lmaster capability
943 when we create the vnnmap. unless there are no nodes available which
944 supports this capability
in which
case we allow the recmaster to
945 become lmaster capable
(temporarily
).
946 - Extend the async framework so that we can use paralell async calls
947 to controls that
return data.
948 - If we
do not have the "can be recmaster" capability
, make sure we will
949 lose any recmaster elections
, unless there are no nodes available that
950 have the capability
, in which
case we "take
/win" the election anyway.
951 - Close and reopen the reclock pnn file at regular intervals.
952 Make it a non
-fatal event
if we occasionally fail to open
/read/write
954 - Monitor that the recovery daemon is still running from the main ctdb
955 daemon and shutdown the main daemon when recovery daemon has terminated.
956 - Add a "ctdb getcapabilities"
command to
read the capabilities off a node.
957 - Define two new capabilities
: can be recmaster and can be lmaster
958 and default both capabilities to YES.
959 - Log denied tcp connection attempts with DEBUG_ERR and not DEBUG_WARNING
960 * Thu Apr
24 2008 : Version 1.0.34
961 - When deleting a public ip from a node
, try to migrate the ip to a different
963 - Change catdb to produce output similar to tdbdump
964 - When adding a new public ip address
, if this ip does not exist yet
in
965 the cluster
, then grab the ip on the
local node and activate it.
966 - When a node disagrees with the recmaster on WHO is the recmaster
, then
967 mark that node as a recovery culprit so it will eventually become
969 - Make ctdb eventscript support the
-n all argument.
970 * Thu Apr
10 2008 : Version 1.0.33
971 - Add facilities to include site
local adaptations to the eventscript
972 by
/etc
/ctdb
/rc.
local which will be
read by all eventscripts.
973 - Add a "ctdb version"
command.
974 - Secure the domain socket with proper permissions from Chris Cowan
975 - Bugfixes
for AIX from Chris Cowan
976 * Wed Apr
02 2008 : Version 1.0.32
977 - Add a control to have a node execute the eventscripts with arbitrary
978 command line arguments.
979 - Add a control "rddumpmemory" that will dump the talloc memory allocations
980 for the recovery daemon.
981 - Decorate the talloc memdump to produce better and easier memory leak
983 - Update the RHEL5 iscsi tgtd scripts to allow one iscsi target
for each
985 - Add two new controls "addip
/delip" that can be used to add
/remove public
986 addresses to a node at runtime. After using these controls a "ctdb recover"
987 ir required to
make the changes take.
988 - Fix a couple of slow memory leaks.
989 * Tue Mar
25 2008 : Version 1.0.31
990 - Add back controls to disable
/enable monitoring on a node.
991 - Fix a memory leak where we used to attach CALL data to the ctdb structure
992 when performing a
local call. Memory which would be lost
if the call was
994 - Reduce the loglevel
for the log output when someone connects to a non
995 public ip address
for samba.
996 - Redo and optimize the vacuuming process to send only one control to each
997 other node containing all records to be vacuumed instead of one
998 control per node per record.
999 * Tue Mar
04 2008 : Version 1.0.30
1000 - Update documentation cor new commands and tuneables
1001 - Add machinereadable output to the ip
,uptime and getdebug commands
1002 - Add a moveip
command to manually failover
/failback public ips
1003 - Add NoIPFallback tuneable that prevents ip address failback
1004 - Use file locking inside the CFS as alternative to verify when other nodes
1005 Are connected
/disconnected to be able to recover from split network
1006 - Add DisableWhenUnhealthy tunable
1007 - Add CTDB_START_AS_DISABLED sysconfig param
1008 - Add
--start
-as
-disabled flag to ctdb
1009 - Add ability to monitor
for OOM condition
1010 * Thu Feb
21 2008 : Version 1.0.29
1011 - Add a new
command to
make expansion of an existing cluster easier
1012 - Fix bug with references to freed objects
in the ctdb structure
1013 - Propagate debuglevel changes to the recovery daemon
1014 - Merge patches to event scripts from Mathieu Parent
:
1015 - MP
: Simulate "service" on systems which
do not provide this tool
1016 - MP
: Set correct permissions
for events.d
/README
1017 - Add nice helper functions to start
/stop nfs from the event scripts
1018 * Fri Feb
08 2008 : Version 1.0.28
1019 - Fix a problem where we tried to use ethtool on non
-ethernet interfaces
1020 - Warn
if the ipvsadm packege is missing when LVS is used
1021 - Dont use absolute pathnames
in some of the event scripts
1022 - Fix
for persistent tdbs growing inifinitely.
1023 * Wed Feb
06 2008 : Version 1.0.27
1024 - Add eventscript
for iscsi
1025 * Thu Jan
31 2008 : Version 1.0.26
1026 - Fix crashbug
in tdb transaction code
1027 * Tue Jan
29 2008 : Version 1.0.25
1028 - added async recovery code
1029 - make event scripts more portable
1030 - fixed ctdb dumpmemory
1031 - more efficient tdb allocation code
1032 - improved machine readable ctdb status output
1034 * Wed Jan
16 2008 : Version 1.0.24
1035 - added syslog support
1036 - documentation updates
1037 * Wed Jan
16 2008 : Version 1.0.23
1038 - fixed a memory leak
in the recoveryd
1039 - fixed a corruption bug
in the new transaction code
1040 - fixed a
case where an packet
for a disconnected client could be processed
1041 - added http event script
1042 - updated documentation
1043 * Thu Jan
10 2008 : Version 1.0.22
1044 - auto
-run vacuum and repack ops
1045 * Wed Jan
09 2008 : Version 1.0.21
1046 - added ctdb vacuum and ctdb repack code
1047 * Sun Jan
06 2008 : Version 1.0.20
1048 - new transaction based recovery code
1049 * Sat Jan
05 2008 : Version 1.0.19
1050 - fixed non
-master bug
1051 - big speedup
in recovery
for large databases
1052 - lots of changes to improve tdb and ctdb
for high churn databases
1053 * Thu Dec
27 2007 : Version 1.0.18
1054 - fixed crash bug
in monitor_handler
1055 * Tue Dec
04 2007 : Version 1.0.17
1056 - fixed bugs related to ban
/unban of nodes
1057 - fixed a race condition that could lead to monitoring being permanently disabled
,
1058 which would lead to long recovery
times
1059 - make deterministic IPs the default
1060 - fixed a bug related to continuous recovery
1061 - added a debugging option
--node
-ip