ctdb-vl.spec 51 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913013113213313413513613713813914014114214314414514614714814915015115215315415515615715815916016116216316416516616716816917017117217317417517617717817918018118218318418518618718818919019119219319419519619719819920020120220320420520620720820921021121221321421521621721821922022122222322422522622722822923023123223323423523623723823924024124224324424524624724824925025125225325425525625725825926026126226326426526626726826927027127227327427527627727827928028128228328428528628728828929029129229329429529629729829930030130230330430530630730830931031131231331431531631731831932032132232332432532632732832933033133233333433533633733833934034134234334434534634734834935035135235335435535635735835936036136236336436536636736836937037137237337437537637737837938038138238338438538638738838939039139239339439539639739839940040140240340440540640740840941041141241341441541641741841942042142242342442542642742842943043143243343443543643743843944044144244344444544644744844945045145245345445545645745845946046146246346446546646746846947047147247347447547647747847948048148248348448548648748848949049149249349449549649749849950050150250350450550650750850951051151251351451551651751851952052152252352452552652752852953053153253353453553653753853954054154254354454554654754854955055155255355455555655755855956056156256356456556656756856957057157257357457557657757857958058158258358458558658758858959059159259359459559659759859960060160260360460560660760860961061161261361461561661761861962062162262362462562662762862963063163263363463563663763863964064164264364464564664764864965065165265365465565665765865966066166266366466566666766866967067167267367467567667767867968068168268368468568668768868969069169269369469569669769869970070170270370470570670770870971071171271371471571671771871972072172272372472572672772872973073173273373473573673773873974074174274374474574674774874975075175275375475575675775875976076176276376476576676776876977077177277377477577677777877978078178278378478578678778878979079179279379479579679779879980080180280380480580680780880981081181281381481581681781881982082182282382482582682782882983083183283383483583683783883984084184284384484584684784884985085185285385485585685785885986086186286386486586686786886987087187287387487587687787887988088188288388488588688788888989089189289389489589689789889990090190290390490590690790890991091191291391491591691791891992092192292392492592692792892993093193293393493593693793893994094194294394494594694794894995095195295395495595695795895996096196296396496596696796896997097197297397497597697797897998098198298398498598698798898999099199299399499599699799899910001001100210031004100510061007100810091010101110121013101410151016101710181019102010211022102310241025102610271028102910301031103210331034103510361037103810391040104110421043104410451046104710481049105010511052105310541055105610571058105910601061106210631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121112211231124112511261127112811291130113111321133113411351136113711381139114011411142114311441145114611471148114911501151115211531154115511561157115811591160116111621163116411651166116711681169
  1. %define initdir %{_sysconfdir}/rc.d/init.d
  2. Summary: A Clustered Database based on Samba's Trivial Database (TDB)
  3. Name: ctdb
  4. Version: 2.0
  5. Release: 1%{?_dist_release}
  6. License: GPLv3+
  7. Group: System Environment/Daemons
  8. URL: http://ctdb.samba.org/
  9. Vendor: Project Vine
  10. Distribution: Vine Linux
  11. Packager: tomop
  12. # Tarfile created using git
  13. # git clone git://git.samba.org/sahlberg/ctdb.git ctdb
  14. # cd ctdb
  15. # git-archive --format=tar --prefix=%{name}-%{version}/ %{name}-%{version} | bzip2 > %{name}-%{version}.tar.bz2
  16. Source0: %{name}-%{version}.tar.gz
  17. # Fedora specific patch, ctdb should not be enabled by default in the runlevels
  18. Patch1: ctdb-no_default_runlevel.patch
  19. Requires: chkconfig coreutils psmisc
  20. Requires: fileutils sed
  21. Requires: tdb-tools
  22. Requires(preun): chkconfig initscripts
  23. Requires(post): chkconfig
  24. Requires(postun): initscripts
  25. BuildRoot: %(mktemp -ud %{_tmppath}/%{name}-%{version}-%{release}-XXXXXX)
  26. BuildRequires: autoconf net-tools popt-devel
  27. %description
  28. CTDB is a cluster implementation of the TDB database used by Samba and other
  29. projects to store temporary data. If an application is already using TDB for
  30. temporary data it is very easy to convert that application to be cluster aware
  31. and use CTDB instead.
  32. %package devel
  33. Group: Development/Libraries
  34. Summary: CTDB clustered database development package
  35. Requires: ctdb = %{version}-%{release}
  36. Provides: ctdb-static = %{version}-%{release}
  37. %description devel
  38. Libraries, include files, etc you can use to develop CTDB applications.
  39. CTDB is a cluster implementation of the TDB database used by Samba and other
  40. projects to store temporary data. If an application is already using TDB for
  41. temporary data it is very easy to convert that application to be cluster aware
  42. and use CTDB instead.
  43. #######################################################################
  44. %prep
  45. %setup -q
  46. # setup the init script and sysconfig file
  47. %setup -T -D -n ctdb-%{version} -q
  48. %patch1 -p1
  49. %build
  50. CC="gcc"
  51. ## always run autogen.sh
  52. ./autogen.sh
  53. CFLAGS="$(echo '%{optflags}') $EXTRA -D_GNU_SOURCE -DCTDB_VERS=\"%{version}-%{release}\"" %configure
  54. make showflags
  55. make %{_smp_mflags}
  56. %install
  57. # Clean up in case there is trash left from a previous build
  58. rm -rf %{buildroot}
  59. # Create the target build directory hierarchy
  60. mkdir -p %{buildroot}%{_sysconfdir}/sysconfig
  61. mkdir -p %{buildroot}%{initdir}
  62. make DESTDIR=%{buildroot} install
  63. install -m644 config/ctdb.sysconfig %{buildroot}%{_sysconfdir}/sysconfig/ctdb
  64. install -m755 config/ctdb.init %{buildroot}%{initdir}/ctdb
  65. mkdir -p %{buildroot}%{_docdir}/ctdb/tests/bin
  66. install -m755 tests/bin/ctdb_transaction %{buildroot}%{_docdir}/ctdb/tests/bin
  67. # Remove "*.old" files
  68. find %{buildroot} -name "*.old" -exec rm -f {} \;
  69. # fix doc path
  70. mv %{buildroot}%{_docdir}/ctdb %{buildroot}%{_docdir}/ctdb-%{version}
  71. cp -r COPYING web %{buildroot}%{_docdir}/ctdb-%{version}
  72. %clean
  73. rm -rf %{buildroot}
  74. %post
  75. /sbin/chkconfig --add ctdb
  76. %preun
  77. if [ "$1" -eq "0" ] ; then
  78. /sbin/service ctdb stop > /dev/null 2>&1
  79. /sbin/chkconfig --del ctdb
  80. fi
  81. %postun
  82. if [ "$1" -ge "1" ]; then
  83. /sbin/service ctdb condrestart >/dev/null 2>&1 || true
  84. fi
  85. # Files section
  86. %files
  87. %defattr(-,root,root,-)
  88. %config(noreplace) %{_sysconfdir}/sysconfig/ctdb
  89. %config(noreplace) %{_sysconfdir}/ctdb/notify.sh
  90. %config(noreplace) %{_sysconfdir}/ctdb/ctdb-crash-cleanup.sh
  91. %config(noreplace) %{_sysconfdir}/ctdb/functions
  92. %attr(755,root,root) %{initdir}/ctdb
  93. %{_docdir}/ctdb-%{version}
  94. %dir %{_sysconfdir}/ctdb
  95. %{_sysconfdir}/ctdb/statd-callout
  96. %{_sysconfdir}/ctdb/events.d/
  97. %{_sysconfdir}/ctdb/debug-hung-script.sh
  98. %{_sysconfdir}/ctdb/gcore_trace.sh
  99. %{_sbindir}/ctdbd
  100. %{_bindir}/ctdb
  101. %{_bindir}/smnotify
  102. %{_bindir}/ping_pong
  103. %{_bindir}/ctdb_diagnostics
  104. %{_bindir}/onnode
  105. %{_bindir}/ltdbtool
  106. %{_mandir}/man1/ctdb.1.gz
  107. %{_mandir}/man1/ctdbd.1.gz
  108. %{_mandir}/man1/onnode.1.gz
  109. %{_mandir}/man1/ltdbtool.1.gz
  110. %{_mandir}/man1/ping_pong.1.gz
  111. %files devel
  112. %defattr(-,root,root,-)
  113. %{_includedir}/ctdb.h
  114. %{_includedir}/ctdb_client.h
  115. %{_includedir}/ctdb_protocol.h
  116. %{_includedir}/ctdb_private.h
  117. %{_includedir}/ctdb_typesafe_cb.h
  118. %{_libdir}/libctdb.a
  119. %{_libdir}/pkgconfig/ctdb.pc
  120. %changelog
  121. * Mon Dec 17 2012 Tomohiro "Tomo-p" KATO <tomop@teamgedoh.net> 2.0-1
  122. - new upstream release.
  123. - initial build for Vine Linux.
  124. * Mon Jun 27 2011 Michael Schwendt <mschwendt@fedoraproject.org> - 1.2.28-2
  125. - Provide virtual -static package to meet guidelines (#700029).
  126. * Mon Apr 18 2011 Sumit Bose <sbose@redhat.com> - 1.2.28-1
  127. - Update to ctdb version 1.2.28
  128. * Tue Feb 08 2011 Fedora Release Engineering <rel-eng@lists.fedoraproject.org> - 1.0.114-2
  129. - Rebuilt for https://fedoraproject.org/wiki/Fedora_15_Mass_Rebuild
  130. * Tue Feb 08 2011 Sumit Bose <sbose@redhat.com> - 1.0.114-1
  131. - Changed $RPM_BUILD_ROOT to %{buildroot}
  132. - Update to ctdb version 1.0.114
  133. - Added patch to fix configure issue
  134. - Added assorted backport patches recommended by upstream developer
  135. * Thu Jan 14 2010 Sumit Bose <sbose@redhat.com> - 1.0.113-1
  136. - Update to ctdb version 1.0.113
  137. * Wed Jan 13 2010 : Version 1.0.113
  138. - Incorrect use of dup2() could cause ctdb to spin eating 100% cpu.
  139. * Tue Jan 12 2010 : Version 1.0.112
  140. - Revert the use of wbinfo --ping-dc as it is proving too unreliable.
  141. - Minor testsuite changes.
  142. * Fri Dec 18 2009 : Version 1.0.111
  143. - Fix a logging bug when an eventscript is aborted that could cause a crash.
  144. - Add back cb_status that was lost in a previous commit.
  145. * Fri Dec 18 2009 : Version 1.0.110
  146. - Metxe: fix for filedescriptor leak in the new eventscript code.
  147. - Rusty: fix for a crash bug in the eventscript code.
  148. * Thu Dec 17 2009 : Version 1.0.109
  149. - Massive eventscript updates. (bz58828)
  150. - Nice the daemon instead of using realtime scheduler, also use mlockall() to
  151. reduce the risk of blockign due to paging.
  152. - Workarounds for valgrind when forking once for each script. Valgrind
  153. consumes massive cpu when terminating the scripts on virtual systems.
  154. - Sync the tdb library with upstream, and use the new TDB_DISALLOW_NESTING
  155. flag.
  156. - Add new command "ctdb dumpdbbackup"
  157. - Start using the new tdb check framework to validate tdb files upon startup.
  158. - A new framework where we can control health for individual tdb databases.
  159. - Fix a crash bug in the logging code.
  160. - New transaction code for persistent databases.
  161. - Various other smaller fixes.
  162. * Tue Dec 8 2009 Sumit Bose <sbose@redhat.com> - 1.0.108-1
  163. - Update to ctdb version 1.0.108
  164. - added fix for bz537223
  165. - added tdb-tools to Requires, fixes bz526479
  166. * Wed Dec 2 2009 Sumit Bose <sbose@redhat.com> - 1.0.107-1
  167. - Update to ctdb version 1.0.107
  168. * Wed Dec 2 2009 : Version 1.0.107
  169. - fix for rusty to solve a double-free that can happen when there are
  170. multiple packets queued and the connection is destroyed before
  171. all packets are processed.
  172. * Tue Dec 1 2009 : Version 1.0.106
  173. - Buildscript changes from Michael Adam
  174. - Dont do a full recovery when there is a mismatch detected for ip addresses,
  175. just do a less disruptive ip-reallocation
  176. - When starting ctdbd, wait until all initial recoveries have finished
  177. before we issue the "startup" event.
  178. So dont start services or monitoring until the cluster has
  179. stabilized.
  180. - Major eventscript overhaul by Ronnie, Rusty and Martins and fixes of a few
  181. bugs found.
  182. * Thu Nov 19 2009 : Version 1.0.105
  183. - Fix a bug where we could SEGV if multiple concurrent "ctdb eventscript ..."
  184. are used and some of them block.
  185. - Monitor the daemon from the syslog child process so we shutdown cleanly when
  186. the main daemon terminates.
  187. - Add a 500k line ringbuffer in memory where all log messages are stored.
  188. - Add a "ctdb getlog <level>" command to pull log messages from the in memory
  189. ringbuffer.
  190. - From martin : fixes to cifs and nfs autotests
  191. - from michael a : fix a bashism in 11.natgw
  192. * Fri Nov 6 2009 : Version 1.0.104
  193. - Suggestion from Metze, we can now use killtcp to kill local connections
  194. for nfs so change the killtcp script to kill both directions of an NFS
  195. connection.
  196. We used to deliberately only kill one direction in these cases due to
  197. limitations.
  198. - Suggestion from christian Ambach, when using natgw, try to avoid using a
  199. UNHEALTHY node as the natgw master.
  200. - From Michael Adam: Fix a SEGV bug in the recent change to the eventscripts
  201. to allow the timeout to apply to each individual script.
  202. - fix a talloc bug in teh vacuuming code that produced nasty valgrind
  203. warnings.
  204. - From Rusty: Set up ulimit to create core files for ctdb, and spawned
  205. processes by default. This is useful for debugging and testing but can be
  206. disabled by setting CTDB_SUPRESS_COREFILE=yes in the sysconfig file.
  207. - Remove the wbinfo -t check from the startup check that winbindd is happy.
  208. - Enhance the test for bond devices so we also check if the sysadmin have
  209. disabled all slave devices using "ifdown".
  210. * Tue Nov 3 2009 : Version 1.0.103
  211. - Dont use vacuuming on persistent databases
  212. - Michael A : transaction updates to persistent databases
  213. - Dont activate service automatically when installing the RPM. Leave this to
  214. the admin.
  215. - Create a child process to send all log messages to, to prevent a hung/slow
  216. syslogd from blocking the main daemon. In this case, discard log messages
  217. instead and let the child process block.
  218. - Michael A: updates to log messages
  219. * Thu Oct 29 2009 : Version 1.0.102
  220. - Wolfgang: fix for the vacuuming code
  221. - Wolfgang: stronger tests for persistent database filename tests
  222. - Improve the log message when we refuse to startup since wbinfo -t fails
  223. to make it easier to spot in the log.
  224. - Update the uptime command output and the man page to indicate that
  225. "time since last ..." if from either the last recovery OR the last failover
  226. - Michael A: transaction updates
  227. * Wed Oct 28 2009 : Version 1.0.101
  228. - create a separate context for non-monitoring events so they dont interfere
  229. with the monitor event
  230. - make sure to return status 0 in teh callback when we abort an event
  231. * Wed Oct 28 2009 : Version 1.0.100
  232. - Change eventscript handling to allow EventScriptTimeout for each individual
  233. script instead of for all scripts as a whole.
  234. - Enhanced logging from the eventscripts, log the name and the duration for
  235. each script as it finishes.
  236. - Add a check to use wbinfo -t for the startup event of samba
  237. - TEMP: allow clients to attach to databases even when teh node is in recovery
  238. mode
  239. - dont run the monitor event as frequently after an event has failed
  240. - DEBUG: in the eventloops, check the local time and warn if the time changes
  241. backward or rapidly forward
  242. - From Metze, fix a bug where recovery master becoming unhealthy did not
  243. trigger an ip failover.
  244. - Disable the multipath script by default
  245. - Automatically re-activate the reclock checking if the reclock file is
  246. specified at runtime. Update manpage to reflect this.
  247. - Add a mechanism where samba can register a SRVID and if samba unexpectedly
  248. disconnects, a message will be broadcasted to all other samba daemons.
  249. - Log the pstree on hung scripts to a file in /tmp isntead of
  250. /var/log/messages
  251. - change ban count before unhealthy/banned to 10
  252. * Thu Oct 22 2009 : Version 1.0.99
  253. - Fix a SEGV in the new db priority code.
  254. - From Wolfgang : eliminate a ctdb_fatal() if there is a dmaster violation
  255. detected.
  256. - During testing we often add/delete eventscripts at runtime. This could cause
  257. an eventscript to fail and mark the node unhealthy if an eventscript was
  258. deleted while we were listing the names. Handle the errorcode and make sure
  259. the node does not becomne unhealthy in this case.
  260. - Lower the debuglevel for the messages when ctdb creates a filedescruiptor so
  261. we dont spam the logs with these messages.
  262. - Dont have the RPM automatically restart ctdb
  263. - Volker : add a missing transaction_cancel() in the handling of persistent
  264. databases
  265. - Treat interfaces with the anme ethX* as bond devices in 10.interfaces so we
  266. do the correct test for if they are up or not.
  267. * Tue Oct 20 2009 : Version 1.0.98
  268. - Fix for the vacuuming database from Wolfgang M
  269. - Create a directory where the test framework can put temporary overrides
  270. to variables and functions.
  271. - Wait a lot longer before shutting down the node when the reclock file
  272. is incorrectly configured, and log where it is configured.
  273. - Try to avoid running the "monitor" event when databases are frozen.
  274. - Add logging for every time we create a filedescriptor so we can trap
  275. fd leaks.
  276. * Thu Oct 14 2009 : Version 1.0.97
  277. - From martins : update onnode.
  278. Update onnode to allow specifying an alternative nodes file from
  279. the command line and also to be able to specify hostnames on the
  280. list of targets :
  281. onnode host1,host2,...
  282. * Wed Oct 14 2009 Sumit Bose <sbose@redhat.com> - 1.0.96-1
  283. - Update to ctdb version 1.0.96
  284. * Tue Oct 13 2009 : Version 1.0.96
  285. - Add more debugging output when eventscripts have trouble. Print a
  286. "pstree -p" to the log when scripts have hung.
  287. - Update the initscript, only print the "No reclock file used" warning
  288. when we do "service ctdb start", dont also print them for all other
  289. actions.
  290. - When changing between unhealthy/healthy state, push a request to the
  291. recovery master to perform an ip reallocation instead of waiting for the
  292. recovery master to pull and check the state change.
  293. - Fix a bug in the new db-priority handling where a pre-.95 recovery master
  294. could no longer lock the databases on a post-.95 daemon.
  295. - Always create the nfs state directories during the "monitor" event.
  296. This makes it easier to configure and enable nfs at runtime.
  297. - From Volker, forward-port a simper deadlock avoiding patch from the 1.0.82
  298. branch. This is a simpler versionof the "db priority lock order" patch
  299. that went into 1.0.95, and will be kept for a few versions until samba
  300. has been updated to use the functionality from 1.0.95.
  301. * Mon Oct 12 2009 : Version 1.0.95
  302. - Add database priorities. Allow samba to set the priority of databases
  303. and lock the databases in priority order during recovery
  304. to avoid a deadlock when samba locks one database then blocks indefinitely
  305. while waiting for the second databaso to become locked.
  306. - Be aggressive and ban nodes where the recovery transaction start call
  307. fails.
  308. * Thu Oct 10 2009 : Version 1.0.94
  309. - Be very aggressive and quickly ban nodes that can not freeze their databases
  310. * Tue Oct 8 2009 : Version 1.0.93
  311. - When adding an ip, make sure to update this assignment on all nodes
  312. so it wont show up as -1 on other nodes.
  313. - When adding an ip and immediately deleting it, it was possible that
  314. the daemon would crash accessing already freed memory.
  315. Readjust the memory hierarchy so the destructors are called in the right
  316. order.
  317. - Add a handshake to the recovery daemon to eliminate some rare cases where
  318. addip/delip might cause a recovery to occur.
  319. - updated onnode documenation from Martin S
  320. - Updates to the natgw eventscript to allow disabling natgw at runtime
  321. * Fri Oct 2 2009 : Version 1.0.92
  322. - Test updates and merge from martin
  323. - Add notification for "startup"
  324. - Add documentation for notification
  325. - from martin, a fix for restarting vsftpd in the eventscript
  326. * Wed Sep 29 2009 Sumit Bose <sbose@redhat.com> - 1.0.91-1
  327. - Update to ctdb version 1.0.91
  328. * Tue Sep 29 2009 : Version 1.0.91
  329. - New vacuum and repack design from Wolgang Mueller.
  330. - Add a new eventscript 01.reclock that will first mark a node unhealthy and
  331. later ban the node if the reclock file can not be accessed.
  332. - Add machinereadable output to the ctdb getreclock command
  333. - merge transaction updates from Michael Adam
  334. - In the new banning code, reset the culprit count to 0 for all nodes that
  335. could successfully compelte a full recovery.
  336. - dont mark the recovery master as a ban culprit because a node in the cluster
  337. needs a recovery. this happens naturally when using ctdb recover command so
  338. dont make this cause a node to be banned.
  339. * Wed Sep 23 2009 Sumit Bose <sbose@redhat.com> - 1.0.90-1
  340. - Update to ctdb version 1.0.90
  341. * Sat Sep 12 2009 : Version 1.0.90
  342. - Be more forgiving for eventscripts that hang during startup
  343. - Fix for a banning bug in the new banning logic
  344. * Thu Sep 3 2009 : Version 1.0.89
  345. - Make it possible to manage winbind independently of samba.
  346. - Add new prototype banning code
  347. - Overwrite the vsftpd state file instead of appending. This eliminates
  348. annoying errors in the log.
  349. - Redirect some iptables commands to dev null
  350. - From Michael A, explicitely set the broadcast when we takeover a public ip
  351. - Remove a reclock file check we no longer need
  352. - Skip any persistent database files ending in .bak
  353. * Mon Aug 17 2009 Sumit Bose <sbose@redhat.com> - 1.0.88-1
  354. - Update to ctdb version 1.0.88
  355. * Mon Aug 17 2009 : Version 1.0.88
  356. - Add a new state for eventscripts : DISABLED.
  357. Add two new commands "ctdb enablescript/disablescript" to enable/disable
  358. eventscripts at runtime.
  359. - Bugfixes for TDB from rusty.
  360. - Merge/Port changes from upstream TDB library by rusty.
  361. - Additional new tests from MartinS. Tests for stop/continue.
  362. - Initial patch to rework vacuuming/repacking process from Wolfgang Mueller.
  363. - Updates from Michael Adam for persistent writes.
  364. - Updates from MartinS to handle the new STOPPED bit in the test framework.
  365. - Make it possible to enable/disable the RECMASTER and LMASTER roles
  366. at runtime. Add two new commands
  367. "ctdb setlmasterrole/setrecmasterrole on/off"
  368. - Make it possible to enable/disable the natgw feature at runtime. Add
  369. the command "ctdb setnatgwstate on/off"
  370. * Fri Jul 24 2009 Fedora Release Engineering <rel-eng@lists.fedoraproject.org> - 1.0.87-2
  371. - Rebuilt for https://fedoraproject.org/wiki/Fedora_12_Mass_Rebuild
  372. * Fri Jul 17 2009 Sumit Bose <sbose@redhat.com> - 1.0.87-1
  373. - Update to ctdb version 1.0.87
  374. * Fri Jul 17 2009 : Version 1.0.87
  375. - Add a new event "stopped" that is called when a node is stopped.
  376. - Documentation of the STOPPED flag and the stop/continue commands
  377. - Make it possible to start a node in STOPPED mode.
  378. - Add a new node flag : STOPPED and commands "ctdb stop" "ctdb continue"
  379. These commands are similar to "diasble/enable" but will also remove the node
  380. from the vnnmap, while disable only fails all ip addresses over.
  381. - tests for NFS , CIFS by martins
  382. - major updates to the init script by martins
  383. - Send gratious arps with a 1.1 second stride instead of a 1 second stride to
  384. workaround interesting "features" of common linux stacks.
  385. - Various test enhancements from martins:
  386. - additional other tests
  387. - add tests for grat arp generation, ping during failover, ssh and failover
  388. - New/updated tcp tickle tests and supprot functions
  389. - provide better debugging when a test fails
  390. - make ctdbd restarts more reliable in the tests
  391. - update the "wait bar" to make the wait progress in tests more obvious
  392. - various cleanups
  393. - when dispatching a message to a handler, make the message a real talloc
  394. object so that we can reparent the object in the tallic hierarchy.
  395. - document the ipreallocate command
  396. - Updates to enable/disable to use the ipreallocate command to block until the
  397. following ipreallocation has completed.
  398. - Update the main daemon and the tools to allow debug level to be a string
  399. instead of an integer.
  400. - Update the sysconfig file to show using string literals instead of numeric
  401. values for the debuglevels used.
  402. - If no debuglevel is specific, make "ctdb setdebug" show the available
  403. options.
  404. - When trying to allocate network packets, add explicit checks if the network
  405. transport has been shutdown before trying and failing, to make log messages
  406. easier to read. Add this extra check and logging to every plave packets are
  407. allocated.
  408. * Wed Jul 1 2009 Sumit Bose <sbose@redhat.com> - 1.0.86-1
  409. - Update to ctdb version 1.0.86
  410. * Tue Jun 30 2009 : Version 1.0.86
  411. - Do not access the reclock at all if VerifyRecoveryLock is zero, not even try
  412. to probe it.
  413. - Allow setting the reclock file as "", which means that no reclock file at
  414. all should be used.
  415. - Document that a reclock file is no longer required, but that it is
  416. dangerous.
  417. - Add a control that can be used to set/clear/change the reclock file in the
  418. daemon during runtime.
  419. - Update the recovery daemon to poll whether a reclock file should be sued and
  420. if so which file at runtime in each monitoring cycle.
  421. - Automatically disable VerifyRecoveryLock everytime a user changes the
  422. location of the reclock file.
  423. - do not allow the VerifyRecoveryLock to be set using ctdb setvar if there is
  424. no recovery lock file specified.
  425. - Add two commands "ctdb getreclock" and "ctdb setreclock" to modify the
  426. reclock file.
  427. * Tue Jun 23 2009 : Version 1.0.85
  428. - From William Jojo : Dont use getopt on AIX
  429. - Make it possible to use "ctdb listnodes" also when the daemon is not running
  430. - Provide machinereadable output to "ctdb listnodes"
  431. - Dont list DELETED nodes in the ctdb listnodes output
  432. - Try to avoid causing a recovery for the average case when
  433. adding/deleting/moving an ip
  434. - When banning a node, drop the IPs on that node only and not all nodes.
  435. - Add tests for NFS and CIFS tickles
  436. - Rename 99.routing to 11.routing so it executes before NFS and LVS scripts
  437. - Increase the default timeout before we deem an unresponsive recovery daemon
  438. hung and shutdown
  439. - Reduce the reclock timout to 5 seconds
  440. - Spawn a child process in the recovery daemon ot check the reclock file to
  441. avoid blocking the process if the underlying filesystem is unresponsive
  442. - fix for filedescriptor leak when a child process timesout
  443. - Dont log errors if waitpid() returns -1
  444. - Onnode updates by Martins
  445. - Test and initscript cleanups from Martin S
  446. * Fri Jun 5 2009 Sumit Bose <sbose@redhat.com> - 1.0.84-1
  447. - Update to ctdb version 1.0.84
  448. * Tue Jun 2 2009 : Version 1.0.84
  449. - Fix a bug in onnode that could not handle dead nodes
  450. * Tue Jun 2 2009 : Version 1.0.83
  451. - Document how to remove a ndoe from a running cluster.
  452. - Hide all deleted nodes from ctdb output.
  453. - Lower the loglevel on some eventscript related items
  454. - Dont queue packets to deleted nodes
  455. - When building initial vnnmap, ignode any nonexisting nodes
  456. - Add a new nodestate : DELETED that is used when deleting a node from an
  457. existing cluster.
  458. - dont remove the ctdb socket when shutting down. This prevents a race in the
  459. initscripts when restarting ctdb quickly after stopping it.
  460. - TDB nesting reworked.
  461. - Remove obsolete ipmux
  462. - From Flavio Carmo Junior: Add eventscript and documentation for ClamAV
  463. antivirus engine
  464. - From Sumit Bose: fix the regex in the test to handle the new ctdb
  465. statistics output that was recently added.
  466. - change the socket type we use for grauitious arps from the obsolete
  467. AF_INET/SOCK_PACKET to instead use PF_PACKET/SOCK_RAW.
  468. - Check return codes for some functions, from Sumit Bose, based on codereview
  469. by Jim Meyering.
  470. - Sumit Bose: Remove structure memeber node_list_file that is no longer used.
  471. - Sumit Bose: fix configure warning for netfilter.h
  472. - Updates to the webpages by Volker.
  473. - Remove error messages about missing /var/log/log.ctdb file from
  474. ctdb_diagnostics.sh from christian Ambach
  475. - Additional error logs if hte eventscript switching from dameon to client
  476. mode fails.
  477. - track how long it takes for ctdbd and the recovery daemon to perform the
  478. rec-lock fcntl() lock attemt and show this in the ctdb statistics output.
  479. * Thu May 14 2009 Sumit Bose <sbose@redhat.com> - 1.0.82-1
  480. - Update to ctdb version 1.0.82
  481. * Thu May 14 2009 : Version 1.0.82
  482. - Update the "ctdb lvsmaster" command to return -1 on error.
  483. - Add a -Y flag to "ctdb lvsmaster"
  484. - RHEL5 apache leaks semaphores when crashing. Add semaphore cleanup to the
  485. 41.httpd eventscript and try to restart apache when it has crashed.
  486. - Fixes to some tests
  487. - Add a -o option to "onnode" which will redirect all stdout to a file for
  488. each of the nodes.
  489. - Add a natgw and a lvs node specifier to onnode so that we can use
  490. "onnode natgw ..."
  491. - Assign the natgw address to lo instead of the private network so it can also
  492. be used where private and public networks are the same.
  493. - Add GPL boilerplates to two missing scripts.
  494. - Change the natgw prefix NATGW_ to CTDB_NATGW_
  495. * Fri May 8 2009 Sumit Bose <sbose@redhat.com> - 1.0.81-1
  496. - Update to ctdb version 1.0.81
  497. * Fri May 8 2009 : Version 1.0.81
  498. - use smbstatus -np instead of smbstatus -n in the 50.samba eventscript
  499. since this avoids performing an expensive traverse on the locking and brlock
  500. databases.
  501. - make ctdb automatically terminate all traverse child processes clusterwide
  502. associated to a client application that terminates before the traversal is
  503. completed.
  504. - From Sumit Bose : fixes to AC_INIT handling.
  505. - From Michael Adam, add Tridge's "ping_pong" tool the the ctdb distro since
  506. this is very useful for testing the backend filesystem.
  507. - From Sumit bose, add support for additional 64 bit platforms.
  508. - Add a link from the webpage to Michael Adams SambaXP paper on CTDB.
  509. * Fri May 1 2009 : Version 1.0.80
  510. - change init shutdown level to 01 for ctdb so it stops before any of the
  511. other services
  512. - if we can not pull a database from a remote node during recovery, mark that
  513. node as a culprit so it becomes banned
  514. - increase the loglevel when we volunteer to drop all ip addresses after
  515. beeing in recovery mode for too long. Make this timeout tuneable with
  516. "RecoveryDropAllIPs" and have it default to 60 seconds
  517. - Add a new flag TDB_NO_NESTING to the tdb layer to prevent nested
  518. transactions which ctdb does not use and does not expect. Have ctdb set this
  519. flag to prevent nested transactions from occuring.
  520. - dont unconditionally kill off ctdb and restrat it on "service ctdb start".
  521. Fail "service ctdb start" with an error if ctdb is already running.
  522. - Add a new tunable "VerifyRecoveryLock" that can be set to 0 to prevent the
  523. main ctdb daemon to verify that the recovery master has locked the reclock
  524. file correctly before allowing it to set the recovery mode to active.
  525. - fix a cosmetic bug with ctdb statistics where certain counters could become
  526. negative.
  527. * Thu Apr 30 2009 Sumit Bose <sbose@redhat.com> - 1.0.79-2
  528. - fixed a ppc64 build issue
  529. * Wed Apr 29 2009 Sumit Bose <sbose@redhat.com> - 1.0.79-1
  530. - Update to ctdb version 1.0.79
  531. * Wed Apr 8 2009 : Version 1.0.79
  532. - From Mathieu Parent: add a ctdb pkgconfig file
  533. - Fix bug 6250
  534. - add a funciton remove_ip to safely remove an ip from an interface, taking
  535. care to workaround an issue with linux alias interfaces.
  536. - Update the natgw eventscript to use the safe remove_ip() function
  537. - fix a bug in the eventscript child process that would cause the socket to be
  538. removed.
  539. - dont verify nodemap on banned nodes during cluster monitoring
  540. - Update the dodgy SeqnumInterval to have ms resolution
  541. * Tue Mar 31 2009 : Version 1.0.78
  542. - Add a notify mechanism so we can send snmptraps/email to external management
  543. systems when the node becomes unhealthy
  544. - include 11.natgw eventscript in thew install so that the NATGW feature works
  545. * Tue Mar 31 2009 : Version 1.0.77
  546. - Update the 99.routing eventscript to also try to add the routes (back)
  547. during a releaseip event. Similar to the reasons why we must add addresses
  548. back during releaseip in 10.interfaces
  549. * Wed Mar 24 2009 : Version 1.0.76
  550. - Add a debugging command "xpnn" which can print the pnn of the node even when
  551. ctdbd is not running.
  552. - Redo the NATGW implementation to allow multiple disjoing NATGW groups in the
  553. same cluster.
  554. * Tue Mar 24 2009 : Version 1.0.75
  555. - Various updates to LVS
  556. - Fix a bug in the killtcp control where we did not set the port correctly
  557. - add a new "ctdb scriptstatus" command that shows the status of the
  558. eventrscripts.
  559. * Mon Mar 16 2009 : Version 1.0.74
  560. - Fixes to AIX from C Cowan.
  561. - Fixes to ctdb_diagnostics so we collect correct GPFS data
  562. - Fixes to the net conf list command in ctdb_diagnostics
  563. - Check the static-routes file IFF it exists in ctdb_diagnostics
  564. * Thu Mar 05 2009 Sumit Bose <sbose@redhat.com> - 1.0.73-1
  565. - Update to ctdb version 1.0.73
  566. * Wed Mar 4 2009 : Version 1.0.73
  567. - Add possibility to disable the check of shares for NFS and Samba
  568. - From Sumit Bose, fix dependencies so make -j works
  569. * Tue Feb 24 2009 Sumit Bose <sbose@redhat.com> - 1.0.72-3
  570. - fix a make -j dependency problem
  571. * Tue Feb 24 2009 Fedora Release Engineering <rel-eng@lists.fedoraproject.org> - 1.0.72-2
  572. - Rebuilt for https://fedoraproject.org/wiki/Fedora_11_Mass_Rebuild
  573. * Wed Feb 18 2009 Sumit Bose <sbose@redhat.com> - 1.0.72-1
  574. - Update to ctdb version 1.0.72
  575. * Wed Feb 18 2009 : Version 1.0.72
  576. - Updates to test scripts by martin s
  577. - Adding a COPYING file
  578. - Use netstat to check for services and ports and fallback to netcat
  579. only if netstat is unavailable.
  580. * Thu Feb 17 2009 Sumit Bose <sbose@redhat.com> - 1.0.71-5
  581. - more fixed according to https://bugzilla.redhat.com/show_bug.cgi?id=459444
  582. * Thu Feb 8 2009 Sumit Bose <sbose@redhat.com> - 1.0.71-4
  583. - added upstream patch with license file
  584. * Thu Feb 6 2009 Sumit Bose <sbose@redhat.com> - 1.0.71-3
  585. - fixed package according to https://bugzilla.redhat.com/show_bug.cgi?id=459444
  586. * Thu Feb 5 2009 Guenther Deschner <gdeschner@redhat.com> - 1.0.71-2
  587. - Update to ctdb version 1.0.71
  588. * Sun Feb 01 2009 : Version 1.0.71
  589. - Additional ipv6 fixes from Michael Adams
  590. * Thu Jan 15 2009 : Version 1.0.70
  591. - IPv6 support is completed. this is backward compatible with ipv4-only
  592. systems. To use IPv6 with samba and ctdb you need current GIT of samba 3.3
  593. or michael adams samba-ctdeb branch.
  594. - Many enhancements to the build system and scripts to make it more SUSE
  595. friendly by Michael Adams.
  596. - Change of how the naming of the package is structured. We are now
  597. using "1.0.70" as a release and "-1" as the revision instead of as
  598. previously using "1.0" as release and ".70" as the revision.
  599. By Michael Adams.
  600. * Wed Dec 17 2008 : Version 1.0.69
  601. - Various fixes to scripts by M Adam
  602. - Dont call ctdb_fatal() when the transport is down during shutdown
  603. * Thu Dec 11 2008 : Version 1.0.68
  604. - Fixes for monitoring of interfaces status from Michael Adam.
  605. - Use -q instead of >/dev/null for grep to enhance readability of the
  606. scripts from Michael Adam.
  607. - Update to the "ctdb recover" command. This command now block until the
  608. has completed. This makes it much easier to use in scripts and avoids
  609. the common workaround :
  610. ctdb recover
  611. ... loop while waiting for recovery completes ...
  612. continue ...
  613. - Add a CTDB_TIMEOUT variable. If set, this variable provides an automatic
  614. timeout for "ctdb <command>", similar to using -T <timeout>
  615. - Set a unique errorcode for "ctdb <command>" when it terminates due to a
  616. timeout so that scripts can distinguish between a hung command and what was
  617. just a failure.
  618. - Update "ctdb ban/unban" so that if the cluster is in recovery these commands
  619. blocks and waits until after recovery is complete before the perform the
  620. ban/unban operation. This is necessary since the recovery process can cause
  621. nodes to become automatically unbanned.
  622. - Update "ctdb ban/unban" to block until the recovery that will follow shortly
  623. after this command has completed.
  624. This makes it much easier to use in scripts and avoids the common
  625. workaround :
  626. ctdb ban/unban
  627. ... loop while waiting for recovery completes ...
  628. continue ...
  629. - Bugfix for the new flags handling in 1.0.67. Abort and restart monitoring
  630. if we failed to get proper nodemaps from a remote node instead of
  631. dereferencing a null pointer.
  632. - If ctdbd was explicitely started with the '--socket' argument, make
  633. ctdbd automatically set CTDB_SOCKET to the specified argument.
  634. This ensures that eventscripts spawned by the ctdb daemon will default to
  635. using the same socket and talk to the correct daemon.
  636. This primarily affects running multiple daemons on the same host and where
  637. you want each instance of ctdb daemons have their eventscripts talk to the
  638. "correct" daemon.
  639. - Update "ctdb ping" to return an error code if the ping fail so that it
  640. can be used in scripts.
  641. - Update to how to synchronize management of node flags across the cluster.
  642. * Tue Dec 02 2008 : Version 1.0.67
  643. - Add a document describing the recovery process.
  644. - Fix a bug in "ctdb setdebug" where it would refuse to set a negative
  645. debug level.
  646. - Print the list of literals for debug names if an invalid one was given
  647. to "ctdb setdebug"
  648. - Redesign how "ctdb reloadnodes" works and reduce the amont of tcp teardowns
  649. used during this event.
  650. - Make it possible to delete a public ip from all nodes at once using
  651. "ctdb delip -n all"
  652. * Sun Nov 23 2008 : Version 1.0.66
  653. - Allow to change the recmaster even when we are not frozen.
  654. - Remove two redundant SAMBA_CHECK variables from the sysconf example
  655. - After a node failure it can take very long before some lock operations
  656. ctdb needs to perform are allowed/works with gpfs again. Workaround this
  657. by treating a hang/timeout as success.
  658. - Dont override CTDB_BASE is fet in the shell already
  659. - Always send keepalive packets regardless of whether the link is idle or not.
  660. - Rewrite the disable/enable flag update logic to prevent a race between
  661. "ctdb disable/enable" and the recovery daemon when updating the flags to
  662. all nodes.
  663. * Wed Nov 12 2008 : Version 1.0.65
  664. - Update the sysconfig example: The default debug level is 2 (NOTICE) and not
  665. 0 (ERROR)
  666. - Add support for a CTDB_SOCKET environment variable for the ctdb command
  667. line tool. If set, this overrides the default socket the ctdb tool will
  668. use.
  669. - Add logging of high latency operations.
  670. * Tue Oct 21 2008 : Version 1.0.64
  671. - Add a context and a timed event so that once we have been in recovery for
  672. too long we drop all public addresses.
  673. * Sun Oct 19 2008 : Version 1.0.63
  674. - Remove logging of "periodic cleanup ..." in 50.samba
  675. - When we reload a nodes file, we must detect this and reload the file also
  676. in the recovery daemon before we try to dereference somethoung beyond the end
  677. of the nodes array.
  678. * Wed Oct 15 2008 : Version 1.0.62
  679. - Allow multiple eventscritps using the same prefix number.
  680. It is undefined which order scripts with the same prefix will execute in.
  681. * Tue Oct 14 2008 : Version 1.0.61
  682. - Use "route add -net" instead of "ip route add" when adding routes in 99.routing
  683. - lower the loglevel os several debug statements
  684. - check the status returned from ctdb_ctrl_get_tickles() before we try to
  685. print them out to the screen.
  686. - install a new eventscript 20.multipathd whoich can be used to monitor that
  687. multipath devices are healthy
  688. * Tue Oct 14 2008 : Version 1.0.60
  689. - Verify that nodes we try to ban/unban are reachable and print an error othervise.
  690. - Update the client and server sides of TAKEIP/RELEASEIP/GETPUBLICIPS and
  691. GETNODEMAP to fall back to the old style ipv4-only controls if the new
  692. ipv4/ipv6 controls fail. This allows an ipv4/v6 enabled ctdb daemon to
  693. interoperate with earlier ipv4-only versions of the daemons.
  694. - From Mathieu Parent : log debian systems log the package versions in ctdb
  695. diagnostics
  696. - From Mathieu Parent : specify logdir location for debian (this patch was
  697. later reversed)
  698. - From Michael Adams : allow # comments in nodes/public_addresses files
  699. * Mon Oct 06 2008 : Version 1.0.59
  700. - Updated "reloadnodes" logic. Instead of bouncing the entire tcp layer it is
  701. sufficient to just close and reopen all outgoing tcp connections.
  702. - New eventscript 99.routing which can be used to re-attach routes to public
  703. interfaces after a takeip event. (routes may be deleted by the kernel when we
  704. release an ip)
  705. - IDR tree fix from Jim Houston
  706. - Better handling of critical events if the local clock is suddenly changed
  707. forward by a lot.
  708. - Fix three slow memory leaks in the recovery daemon
  709. - New ctdb command : ctdb recmaster which prints the pnn of the recmaster
  710. - Onnode enhancements from Martin S : "healthy" and "connected" are now
  711. possible nodespecifiers
  712. - From Martin S : doc fixes
  713. - lowering some debug levels for some nonvital informational messages
  714. - Make the daemon daemon monitoring stronger and allow ctdbd to detect a hung
  715. recovery daemon.
  716. - From C Cowan : patches to compile ipv6 under AIX
  717. - zero out some structs to keep valgrind happy
  718. * Mon Sep 8 2008 Abhijith Das <adas@redhat.com> - 1.0.58-1
  719. - This release repackages upstream's version 1.0.58 for fedora
  720. * Wed Aug 27 2008 : Version 1.0.58
  721. - revert the name change tcp_tcp_client back to tcp_control_tcp so
  722. samba can build.
  723. - Updates to the init script from Abhijith Das <adas@redhat.com>
  724. * Mon Aug 25 2008 : Version 1.0.57
  725. - initial support for IPv6
  726. * Mon Aug 11 2008 : Version 1.0.56
  727. - fix a memory leak in the recovery daemon.
  728. * Mon Aug 11 2008 : Version 1.0.55
  729. - Fix the releaseip message we seond to samba.
  730. * Fri Aug 8 2008 : Version 1.0.54
  731. - fix a looping error in the transaction code
  732. - provide a more detailed error code for persistent store errors
  733. so clients can make more intelligent choices on how to try to recover
  734. * Thu Aug 7 2008 : Version 1.0.53
  735. - Remove the reclock.pnn file it can cause gpfs to fail to umount
  736. - New transaction code
  737. * Mon Aug 4 2008 : Version 1.0.52
  738. - Send an explicit gratious arp when starting sending the tcp tickles.
  739. - When doing failover, issue a killtcp to non-NFS/non-CIFS clients
  740. so that they fail quickly. NFS and CIFS already fail and recover
  741. quickly.
  742. - Update the test scripts to handle CTRL-C to kill off the test.
  743. * Mon Jul 28 2008 : Version 1.0.51
  744. - Strip off the vlan tag from bond devices before we check in /proc
  745. if the interface is up or not.
  746. - Use testparm in the background in the scripts to allow probing
  747. that the shares do exist.
  748. - Fix a bug in the logging code to handle multiline entries better
  749. - Rename private elements from private to private_data
  750. * Fri Jul 18 2008 : Version 1.0.50
  751. - Dont assume that just because we can establish a TCP connection
  752. that we are actually talking to a functioning ctdb daemon.
  753. So dont mark the node as CONNECTED just because the tcp handshake
  754. was successful.
  755. - Dont try to set the recmaster to ourself during elections for those
  756. cases we know this will fail. To remove some annoying benign but scary
  757. looking entries from the log.
  758. - Bugfix for eventsystem for signal handling that could cause a node to
  759. hang.
  760. * Thu Jul 17 2008 : Version 1.0.49
  761. - Update the safe persistent update fix to work with unpatched samba
  762. servers.
  763. * Thu Jul 17 2008 : Version 1.0.48
  764. - Update the spec file.
  765. - Do not start new user-triggered eventscripts if we are already
  766. inside recovery mode.
  767. - Add two new controls to start/cancel a persistent update.
  768. A client such as samba can use these to tell ctdbd that it will soon
  769. be writing directly to the persistent database tdb file. So if
  770. samba is -9ed before it has eitehr done the persistent_store or
  771. canceled the operation, ctdb knows that the persistent databases
  772. 'may' be out of sync and therefore a full blown recovery is called for.
  773. - Add two new options :
  774. CTDB_SAMBA_SKIP_CONF_CHECK and CTDB_SAMBA_CHECK_PORTS that can be used
  775. to override what checks to do when monitoring samba health.
  776. We can no longer use the smbstatus, net or testparm commands to check
  777. if samba or its config is healthy since these commands may block
  778. indefinitely and thus can not be used in scripts.
  779. * Fri Jul 11 2008 : Version 1.0.47
  780. - Fix a double free bug where if a user striggered (ctdb eventscript)
  781. hung and while the timeout handler was being processed a new user
  782. triggered eventscript was started we would free state twice.
  783. - Rewrite of onnode and associated documentation.
  784. * Thu Jul 10 2008 : Version 1.0.46
  785. - Document both the LVS:cingle-ip-address and the REMOTE-NODE:wan-accelerator
  786. capabilities.
  787. - Add commands "ctdb pnn", "ctdb lvs", "ctdb lvsmaster".
  788. - LVS improvements. LVS is the single-ip-address mode for a ctdb cluster.
  789. - Fixes to supress rpmlint warnings
  790. - AXI compile fixes.
  791. - Change \s to [[:space:]] in some scripts. Not all RHEL5 packages come
  792. with a egrep that handles \s even same version but different arch.
  793. - Revert the change to NFS restart. CTDB should NOT attempt to restart
  794. failed services.
  795. - Rewrite of the waitpid() patch to use the eventsystem for handling
  796. signals.
  797. * Tue Jul 8 2008 : Version 1.0.45
  798. - Try to restart the nfs service if it has failed to respond 3 times in a row.
  799. - waitpid() can block if the child does not respond promptly to SIGTERM.
  800. ignore all SIGCHILD signals by setting SIGCHLD to SIG_DEF.
  801. get rid of all calls to waitpid().
  802. - make handling of eventscripts hanging more liberal.
  803. only consider the script to have failed and making the node unhealthy
  804. IF the eventscript terminated wiht an error
  805. OR the eventscript hung 5 or more times in a row
  806. * Mon Jul 7 2008 : Version 1.0.44
  807. - Add a CTDB_VALGRIND option to /etc/sysconfig/ctdb to make it start
  808. ctdb under valgrind. Logs go to /var/log/ctdb_valgrind.PID
  809. - Add a hack to show the control opcode that caused uninitialized data
  810. in the valgrind output by encoding the opcode as the line number.
  811. - Initialize structures and allocated memory in various places in
  812. ctdb to make it valgrind-clean and remove all valgrind errors/warnings.
  813. - If/when we destroy a lockwait child, also make sure we cancel any pending transactions
  814. - If a transaction_commit fails, delete/cancel any pending transactions and
  815. return an error instead of calling ctdb_fatal()
  816. - When running ctdb under valgrind, make sure we run it with --nosetsched and also
  817. ensure that we do not use mem-mapped i/o when accessing the tdb's.
  818. - zero out ctdb->freeze_handle when we free/destroy a freeze-child.
  819. This prevent a heap corruption/ctdb crash bug that could trigger
  820. if the freeze child times out.
  821. - we dont need to explicitely thaw the databases from the recovery daemon
  822. since this is done implicitely when we restore the recovery mode back to normal.
  823. - track when we start and stop a recovery. Add the 'time it took to complete the
  824. recovery' to the 'ctdb uptime' output.
  825. Ensure by tracking the start/stop recovery timestamps that we do not
  826. check that the ip allocation is consistend from inside the recovery daemon
  827. while a different node (recovery master) is performing a recovery.
  828. This prevent a race that could cause a full recovery to trigger if the
  829. 'ctdb disable/enable' commands took very long.
  830. - The freeze child indicates to the master daemon that all databases are locked
  831. by writing data to the pipe shared with the master daemon.
  832. This write sometimes fail and thus the master daemon never notices that the databases
  833. are locked cvausing long timeouts and extra recoveries.
  834. Check that the write is successful and try the write again if it failed.
  835. - In each node, verify that the recmaster have the right node flags for us
  836. and force a push of our flags to the recmaster if wrong.
  837. * Tue Jul 1 2008 : Version 1.0.43
  838. - Updates and bugfixes to the specfile to keep rpmlint happy
  839. - Force a global flags update after each recovery event.
  840. - Verify that the recmaster agrees with our node flags and update the
  841. recmaster othervise.
  842. - When writing back to the parent from a freeze-child across the pipe,
  843. loop over the write in case the write failed with an error othervise
  844. the parent will never be notified tha the child has completed the operation.
  845. - Automatically thaw all databases when recmaster marks us as being in normal
  846. mode instead of recovery mode.
  847. * Fri Jun 13 2008 : Version 1.0.42
  848. - When event scripts have hung/timedout more than EventScriptBanCount times
  849. in a row the node will ban itself.
  850. - Many updates to persistent write tests and the test scripts.
  851. * Wed May 28 2008 : Version 1.0.41
  852. - Reactivate the safe writes to persistent databases and solve the
  853. locking issues. Locking issues are solved the only possible way,
  854. by using a child process to do the writes. Expensive and slow but... .
  855. * Tue May 27 2008 : Version 1.0.40
  856. - Read the samba sysconfig file from the 50.samba eventscript
  857. - Fix some emmory hierarchical bugs in the persistent write handling
  858. * Thu May 22 2008 : Version 1.0.39
  859. - Moved a CTDB_MANAGES_NFS, CTDB_MANAGES_ISCSI and CTDB_MANAGES_CSFTPD
  860. into /etc/sysconfig/ctdb
  861. - Lowered some debug messages to not fill the logfile with entries
  862. that normally occur in the default configuration.
  863. * Fri May 16 2008 : Version 1.0.38
  864. - Add machine readable output support to "ctdb getmonmode"
  865. - Lots of tweaks and enhancements if the event scripts are "slow"
  866. - Merge from tridge: an attempt to break the chicken-and-egg deadlock that
  867. net conf introduces if used from an eventscript.
  868. - Enhance tickles so we can tickle an ipv6 connection.
  869. - Start adding ipv6 support : create a new container to replace sockaddr_in.
  870. - Add a checksum routine for ipv6/tcp
  871. - When starting up ctdb, let the init script do a tdbdump on all
  872. persistent databases and verify that they are good (i.e. not corrupted).
  873. - Try to use "safe transactions" when writing to a persistent database
  874. that was opened with the TDB_NOSYNC flag. If we can get the transaction
  875. thats great, if we cant we have to write anyway since we cant block here.
  876. * Mon May 12 2008 : Version 1.0.37
  877. - When we shutdown ctdb we close the transport down before we run the
  878. "shutdown" eventscripts. If ctdb decides to send a packet to a remote node
  879. after we have shutdown the transport but before we have shutdown ctdbd
  880. itself this could lead to a SEGV instead of a clean shutdown. Fix.
  881. - When using the "exportfs" command to extract which NFS export directories
  882. to monitor, exportfs violates the "principle of least surprise" and
  883. sometimes report a single export line as two lines of text output
  884. causing the monitoring to fail.
  885. * Fri May 9 2008 : Version 1.0.36
  886. - fix a memory corruption bug that could cause the recovery daemon to crash.
  887. - fix a bug with distributing public ip addresses during recovery.
  888. If the node that is the recovery master did NOT use public addresses,
  889. then it assumed that no other node in the cluster used them either and
  890. thus skipped the entire step of reallocating public addresses.
  891. * Wed May 7 2008 : Version 1.0.35
  892. - During recovery, when we define the new set of lmasters (vnnmap)
  893. only consider those nodes that have the can-be-lmaster capability
  894. when we create the vnnmap. unless there are no nodes available which
  895. supports this capability in which case we allow the recmaster to
  896. become lmaster capable (temporarily).
  897. - Extend the async framework so that we can use paralell async calls
  898. to controls that return data.
  899. - If we do not have the "can be recmaster" capability, make sure we will
  900. lose any recmaster elections, unless there are no nodes available that
  901. have the capability, in which case we "take/win" the election anyway.
  902. - Close and reopen the reclock pnn file at regular intervals.
  903. Make it a non-fatal event if we occasionally fail to open/read/write
  904. to this file.
  905. - Monitor that the recovery daemon is still running from the main ctdb
  906. daemon and shutdown the main daemon when recovery daemon has terminated.
  907. - Add a "ctdb getcapabilities" command to read the capabilities off a node.
  908. - Define two new capabilities : can be recmaster and can be lmaster
  909. and default both capabilities to YES.
  910. - Log denied tcp connection attempts with DEBUG_ERR and not DEBUG_WARNING
  911. * Thu Apr 24 2008 : Version 1.0.34
  912. - When deleting a public ip from a node, try to migrate the ip to a different
  913. node first.
  914. - Change catdb to produce output similar to tdbdump
  915. - When adding a new public ip address, if this ip does not exist yet in
  916. the cluster, then grab the ip on the local node and activate it.
  917. - When a node disagrees with the recmaster on WHO is the recmaster, then
  918. mark that node as a recovery culprit so it will eventually become
  919. banned.
  920. - Make ctdb eventscript support the -n all argument.
  921. * Thu Apr 10 2008 : Version 1.0.33
  922. - Add facilities to include site local adaptations to the eventscript
  923. by /etc/ctdb/rc.local which will be read by all eventscripts.
  924. - Add a "ctdb version" command.
  925. - Secure the domain socket with proper permissions from Chris Cowan
  926. - Bugfixes for AIX from Chris Cowan
  927. * Wed Apr 02 2008 : Version 1.0.32
  928. - Add a control to have a node execute the eventscripts with arbitrary
  929. command line arguments.
  930. - Add a control "rddumpmemory" that will dump the talloc memory allocations
  931. for the recovery daemon.
  932. - Decorate the talloc memdump to produce better and easier memory leak
  933. tracking.
  934. - Update the RHEL5 iscsi tgtd scripts to allow one iscsi target for each
  935. public address.
  936. - Add two new controls "addip/delip" that can be used to add/remove public
  937. addresses to a node at runtime. After using these controls a "ctdb recover"
  938. ir required to make the changes take.
  939. - Fix a couple of slow memory leaks.
  940. * Tue Mar 25 2008 : Version 1.0.31
  941. - Add back controls to disable/enable monitoring on a node.
  942. - Fix a memory leak where we used to attach CALL data to the ctdb structure
  943. when performing a local call. Memory which would be lost if the call was
  944. aborted.
  945. - Reduce the loglevel for the log output when someone connects to a non
  946. public ip address for samba.
  947. - Redo and optimize the vacuuming process to send only one control to each
  948. other node containing all records to be vacuumed instead of one
  949. control per node per record.
  950. * Tue Mar 04 2008 : Version 1.0.30
  951. - Update documentation cor new commands and tuneables
  952. - Add machinereadable output to the ip,uptime and getdebug commands
  953. - Add a moveip command to manually failover/failback public ips
  954. - Add NoIPFallback tuneable that prevents ip address failback
  955. - Use file locking inside the CFS as alternative to verify when other nodes
  956. Are connected/disconnected to be able to recover from split network
  957. - Add DisableWhenUnhealthy tunable
  958. - Add CTDB_START_AS_DISABLED sysconfig param
  959. - Add --start-as-disabled flag to ctdb
  960. - Add ability to monitor for OOM condition
  961. * Thu Feb 21 2008 : Version 1.0.29
  962. - Add a new command to make expansion of an existing cluster easier
  963. - Fix bug with references to freed objects in the ctdb structure
  964. - Propagate debuglevel changes to the recovery daemon
  965. - Merge patches to event scripts from Mathieu Parent :
  966. - MP: Simulate "service" on systems which do not provide this tool
  967. - MP: Set correct permissions for events.d/README
  968. - Add nice helper functions to start/stop nfs from the event scripts
  969. * Fri Feb 08 2008 : Version 1.0.28
  970. - Fix a problem where we tried to use ethtool on non-ethernet interfaces
  971. - Warn if the ipvsadm packege is missing when LVS is used
  972. - Dont use absolute pathnames in some of the event scripts
  973. - Fix for persistent tdbs growing inifinitely.
  974. * Wed Feb 06 2008 : Version 1.0.27
  975. - Add eventscript for iscsi
  976. * Thu Jan 31 2008 : Version 1.0.26
  977. - Fix crashbug in tdb transaction code
  978. * Tue Jan 29 2008 : Version 1.0.25
  979. - added async recovery code
  980. - make event scripts more portable
  981. - fixed ctdb dumpmemory
  982. - more efficient tdb allocation code
  983. - improved machine readable ctdb status output
  984. - added ctdb uptime
  985. * Wed Jan 16 2008 : Version 1.0.24
  986. - added syslog support
  987. - documentation updates
  988. * Wed Jan 16 2008 : Version 1.0.23
  989. - fixed a memory leak in the recoveryd
  990. - fixed a corruption bug in the new transaction code
  991. - fixed a case where an packet for a disconnected client could be processed
  992. - added http event script
  993. - updated documentation
  994. * Thu Jan 10 2008 : Version 1.0.22
  995. - auto-run vacuum and repack ops
  996. * Wed Jan 09 2008 : Version 1.0.21
  997. - added ctdb vacuum and ctdb repack code
  998. * Sun Jan 06 2008 : Version 1.0.20
  999. - new transaction based recovery code
  1000. * Sat Jan 05 2008 : Version 1.0.19
  1001. - fixed non-master bug
  1002. - big speedup in recovery for large databases
  1003. - lots of changes to improve tdb and ctdb for high churn databases
  1004. * Thu Dec 27 2007 : Version 1.0.18
  1005. - fixed crash bug in monitor_handler
  1006. * Tue Dec 04 2007 : Version 1.0.17
  1007. - fixed bugs related to ban/unban of nodes
  1008. - fixed a race condition that could lead to monitoring being permanently disabled,
  1009. which would lead to long recovery times
  1010. - make deterministic IPs the default
  1011. - fixed a bug related to continuous recovery
  1012. - added a debugging option --node-ip