1CTDB(1) CTDB - clustered TDB database CTDB(1)
2
6 ctdb - CTDB management utility
7
9 ctdb [OPTION...] {COMMAND} [COMMAND-ARGS]
10
12 ctdb is a utility to view and manage a CTDB cluster.
13 The following terms are used when referring to nodes in a cluster:
15 PNN
17 Physical Node Number. The physical node number is an integer that
18 describes the node in the cluster. The first node has physical node
19 number 0. in a cluster.
20 PNN-LIST
22 This is either a single PNN, a comma-separate list of PNNs or
23 "all".
24 Commands that reference a database use the following terms:
26 DB
28 This is either a database name, such as locking.tdb or a database
29 ID such as "0x42fe72c5".
30 DB-LIST
32 A space separated list of at least one DB.
33
35 -n PNN
36 The node specified by PNN should be queried for the requested
37 information. Default is to query the daemon running on the local
38 host.
39 -Y
41 Produce output in machine readable form for easier parsing by
42 scripts. This uses a field delimiter of ':'. Not all commands
43 support this option.
44 -x SEPARATOR
46 Use SEPARATOR to delimit fields in machine readable output. This
47 implies -Y.
48 -X
50 Produce output in machine readable form for easier parsing by
51 scripts. This uses a field delimiter of '|'. Not all commands
52 support this option.
53 This is equivalent to "-x|" and avoids some shell quoting issues.
55 -t TIMEOUT
57 Indicates that ctdb should wait up to TIMEOUT seconds for a
58 response to most commands sent to the CTDB daemon. The default is
59 10 seconds.
60 -T TIMELIMIT
62 Indicates that TIMELIMIT is the maximum run time (in seconds) for
63 the ctdb command. When TIMELIMIT is exceeded the ctdb command will
64 terminate with an error. The default is 120 seconds.
65 -? --help
67 Print some help text to the screen.
68 --usage
70 Print useage information to the screen.
71 -d --debug=DEBUGLEVEL
73 Change the debug level for the command. Default is NOTICE.
74
76 These are commands used to monitor and administer a CTDB cluster.
77 pnn
79 This command displays the PNN of the current node.
80 status
82 This command shows the current status of all CTDB nodes based on
83 information from the queried node.
84 Note: If the the queried node is INACTIVE then the status might not be
86 current.
87 Node status
89 This includes the number of physical nodes and the status of each
90 node. See ctdb(7) for information about node states.
91 Generation
93 The generation id is a number that indicates the current generation
94 of a cluster instance. Each time a cluster goes through a
95 reconfiguration or a recovery its generation id will be changed.
96 This number does not have any particular meaning other than to keep
98 track of when a cluster has gone through a recovery. It is a random
99 number that represents the current instance of a ctdb cluster and
100 its databases. The CTDB daemon uses this number internally to be
101 able to tell when commands to operate on the cluster and the
102 databases was issued in a different generation of the cluster, to
103 ensure that commands that operate on the databases will not survive
104 across a cluster database recovery. After a recovery, all old
105 outstanding commands will automatically become invalid.
106 Sometimes this number will be shown as "INVALID". This only means
108 that the ctdbd daemon has started but it has not yet merged with
109 the cluster through a recovery. All nodes start with generation
110 "INVALID" and are not assigned a real generation id until they have
111 successfully been merged with a cluster through a recovery.
112 Virtual Node Number (VNN) map
114 Consists of the number of virtual nodes and mapping from virtual
115 node numbers to physical node numbers. Only nodes that are
116 participating in the VNN map can become lmaster for database
117 records.
118 Recovery mode
120 This is the current recovery mode of the cluster. There are two
121 possible modes:
122 NORMAL - The cluster is fully operational.
124 RECOVERY - The cluster databases have all been frozen, pausing all
126 services while the cluster awaits a recovery process to complete. A
127 recovery process should finish within seconds. If a cluster is
128 stuck in the RECOVERY state this would indicate a cluster
129 malfunction which needs to be investigated.
130 Once the recovery master detects an inconsistency, for example a
132 node becomes disconnected/connected, the recovery daemon will
133 trigger a cluster recovery process, where all databases are
134 remerged across the cluster. When this process starts, the recovery
135 master will first "freeze" all databases to prevent applications
136 such as samba from accessing the databases and it will also mark
137 the recovery mode as RECOVERY.
138 When the CTDB daemon starts up, it will start in RECOVERY mode.
140 Once the node has been merged into a cluster and all databases have
141 been recovered, the node mode will change into NORMAL mode and the
142 databases will be "thawed", allowing samba to access the databases
143 again.
144 Recovery master
146 This is the cluster node that is currently designated as the
147 recovery master. This node is responsible of monitoring the
148 consistency of the cluster and to perform the actual recovery
149 process when reqired.
150 Only one node at a time can be the designated recovery master.
152 Which node is designated the recovery master is decided by an
153 election process in the recovery daemons running on each node.
154 Example
156 # ctdb status
157 Number of nodes:4
158 pnn:0 192.168.2.200 OK (THIS NODE)
159 pnn:1 192.168.2.201 OK
160 pnn:2 192.168.2.202 OK
161 pnn:3 192.168.2.203 OK
162 Generation:1362079228
163 Size:4
164 hash:0 lmaster:0
165 hash:1 lmaster:1
166 hash:2 lmaster:2
167 hash:3 lmaster:3
168 Recovery mode:NORMAL (0)
169 Recovery master:0
170 nodestatus [PNN-LIST]
173 This command is similar to the status command. It displays the "node
174 status" subset of output. The main differences are:
175 · The exit code is the bitwise-OR of the flags for each specified
177 node, while ctdb status exits with 0 if it was able to retrieve
178 status for all nodes.
179 · ctdb status provides status information for all nodes. ctdb
181 nodestatus defaults to providing status for only the current node.
182 If PNN-LIST is provided then status is given for the indicated
183 node(s).
184 A common invocation in scripts is ctdb nodestatus all to check whether
186 all nodes in a cluster are healthy.
187 Example
189 # ctdb nodestatus
190 pnn:0 10.0.0.30 OK (THIS NODE)
191 # ctdb nodestatus all
193 Number of nodes:2
194 pnn:0 10.0.0.30 OK (THIS NODE)
195 pnn:1 10.0.0.31 OK
196 recmaster
199 This command shows the pnn of the node which is currently the
200 recmaster.
201 Note: If the the queried node is INACTIVE then the status might not be
203 current.
204 uptime
206 This command shows the uptime for the ctdb daemon. When the last
207 recovery or ip-failover completed and how long it took. If the
208 "duration" is shown as a negative number, this indicates that there is
209 a recovery/failover in progress and it started that many seconds ago.
210 Example
212 # ctdb uptime
213 Current time of node : Thu Oct 29 10:38:54 2009
214 Ctdbd start time : (000 16:54:28) Wed Oct 28 17:44:26 2009
215 Time of last recovery/failover: (000 16:53:31) Wed Oct 28 17:45:23 2009
216 Duration of last recovery/failover: 2.248552 seconds
217 listnodes
220 This command shows lists the ip addresses of all the nodes in the
221 cluster.
222 Example
224 # ctdb listnodes
225 192.168.2.200
226 192.168.2.201
227 192.168.2.202
228 192.168.2.203
229 natgw {master|list|status}
232 This command shows different aspects of NAT gateway status. For an
233 overview of CTDB's NAT gateway functionality please see the NAT GATEWAY
234 section in ctdb(7).
235 master
237 Show the PNN and private IP address of the current NAT gateway
238 master node.
239 Example output:
241 1 192.168.2.201
243 list
246 List the private IP addresses of nodes in the current NAT gateway
247 group, annotating the master node.
248 Example output:
250 192.168.2.200
252 192.168.2.201 MASTER
253 192.168.2.202
254 192.168.2.203
255 status
258 List the nodes in the current NAT gateway group and their status.
259 Example output:
261 pnn:0 192.168.2.200 UNHEALTHY (THIS NODE)
263 pnn:1 192.168.2.201 OK
264 pnn:2 192.168.2.202 OK
265 pnn:3 192.168.2.203 OK
266 ping
269 This command will "ping" specified CTDB nodes in the cluster to verify
270 that they are running.
271 Example
273 # ctdb ping
274 response from 0 time=0.000054 sec (3 clients)
275 ifaces
278 This command will display the list of network interfaces, which could
279 host public addresses, along with their status.
280 Example
282 # ctdb ifaces
283 Interfaces on node 0
284 name:eth5 link:up references:2
285 name:eth4 link:down references:0
286 name:eth3 link:up references:1
287 name:eth2 link:up references:1
288 # ctdb -X ifaces
290 |Name|LinkStatus|References|
291 |eth5|1|2|
292 |eth4|0|0|
293 |eth3|1|1|
294 |eth2|1|1|
295 ip
298 This command will display the list of public addresses that are
299 provided by the cluster and which physical node is currently serving
300 this ip. By default this command will ONLY show those public addresses
301 that are known to the node itself. To see the full list of all public
302 ips across the cluster you must use "ctdb ip all".
303 Example
305 # ctdb ip -v
306 Public IPs on node 0
307 172.31.91.82 node[1] active[] available[eth2,eth3] configured[eth2,eth3]
308 172.31.91.83 node[0] active[eth3] available[eth2,eth3] configured[eth2,eth3]
309 172.31.91.84 node[1] active[] available[eth2,eth3] configured[eth2,eth3]
310 172.31.91.85 node[0] active[eth2] available[eth2,eth3] configured[eth2,eth3]
311 172.31.92.82 node[1] active[] available[eth5] configured[eth4,eth5]
312 172.31.92.83 node[0] active[eth5] available[eth5] configured[eth4,eth5]
313 172.31.92.84 node[1] active[] available[eth5] configured[eth4,eth5]
314 172.31.92.85 node[0] active[eth5] available[eth5] configured[eth4,eth5]
315 # ctdb -X ip -v
317 |Public IP|Node|ActiveInterface|AvailableInterfaces|ConfiguredInterfaces|
318 |172.31.91.82|1||eth2,eth3|eth2,eth3|
319 |172.31.91.83|0|eth3|eth2,eth3|eth2,eth3|
320 |172.31.91.84|1||eth2,eth3|eth2,eth3|
321 |172.31.91.85|0|eth2|eth2,eth3|eth2,eth3|
322 |172.31.92.82|1||eth5|eth4,eth5|
323 |172.31.92.83|0|eth5|eth5|eth4,eth5|
324 |172.31.92.84|1||eth5|eth4,eth5|
325 |172.31.92.85|0|eth5|eth5|eth4,eth5|
326 ipinfo IP
329 This command will display details about the specified public addresses.
330 Example
332 # ctdb ipinfo 172.31.92.85
333 Public IP[172.31.92.85] info on node 0
334 IP:172.31.92.85
335 CurrentNode:0
336 NumInterfaces:2
337 Interface[1]: Name:eth4 Link:down References:0
338 Interface[2]: Name:eth5 Link:up References:2 (active)
339 event run|status|script list|script enable|script disable
342 This command is used to control event daemon and to inspect status of
343 various events.
344 The commands below require a component to be specified. In the current
346 version the only valid component is legacy.
347 run TIMEOUT COMPONENT EVENT [ARGUMENTS]
349 This command can be used to manually run specified EVENT in
350 COMPONENT with optional ARGUMENTS. The event will be allowed to run
351 a maximum of TIMEOUT seconds. If TIMEOUT is 0, then there is no
352 time limit for running the event.
353 status COMPONENT EVENT
355 This command displays the last execution status of the specified
356 EVENT in COMPONENT.
357 The command will terminate with the exit status corresponding to
359 the overall status of event that is displayed.
360 The output is the list of event scripts executed. Each line shows
362 the name, status, duration and start time for each script.
363 Example output:
365 00.ctdb OK 0.014 Sat Dec 17 19:39:11 2016
367 01.reclock OK 0.013 Sat Dec 17 19:39:11 2016
368 05.system OK 0.029 Sat Dec 17 19:39:11 2016
369 06.nfs OK 0.014 Sat Dec 17 19:39:11 2016
370 10.interface OK 0.037 Sat Dec 17 19:39:11 2016
371 11.natgw OK 0.011 Sat Dec 17 19:39:11 2016
372 11.routing OK 0.007 Sat Dec 17 19:39:11 2016
373 13.per_ip_routing OK 0.007 Sat Dec 17 19:39:11 2016
374 20.multipathd OK 0.007 Sat Dec 17 19:39:11 2016
375 31.clamd OK 0.007 Sat Dec 17 19:39:11 2016
376 40.vsftpd OK 0.013 Sat Dec 17 19:39:11 2016
377 41.httpd OK 0.018 Sat Dec 17 19:39:11 2016
378 49.winbind OK 0.023 Sat Dec 17 19:39:11 2016
379 50.samba OK 0.100 Sat Dec 17 19:39:12 2016
380 60.nfs OK 0.376 Sat Dec 17 19:39:12 2016
381 70.iscsi OK 0.009 Sat Dec 17 19:39:12 2016
382 91.lvs OK 0.007 Sat Dec 17 19:39:12 2016
383 script list COMPONENT
386 List the available event scripts in COMPONENT. Enabled scripts are
387 flagged with a '*'.
388 Generally, event scripts are provided by CTDB. However, local or
390 3rd party event scripts may also be available. These are shown in a
391 separate section after those provided by CTDB.
392 Example output:
394 * 00.ctdb
396 * 01.reclock
397 * 05.system
398 * 06.nfs
399 * 10.interface
400 11.natgw
401 11.routing
402 13.per_ip_routing
403 20.multipathd
404 31.clamd
405 40.vsftpd
406 41.httpd
407 * 49.winbind
408 * 50.samba
409 * 60.nfs
410 70.iscsi
411 91.lvs
412 * 02.local
414 script enable COMPONENT SCRIPT
417 Enable the specified event SCRIPT in COMPONENT. Only enabled
418 scripts will be executed when running any event.
419 script disable COMPONENT SCRIPT
421 Disable the specified event SCRIPT in COMPONENT. This will prevent
422 the script from executing when running any event.
423 scriptstatus
425 This command displays which event scripts where run in the previous
426 monitoring cycle and the result of each script. If a script failed with
427 an error, causing the node to become unhealthy, the output from that
428 script is also shown.
429 This command is deprecated. It's provided for backward compatibility.
431 In place of ctdb scriptstatus, use ctdb event status.
432 Example
434 # ctdb scriptstatus
435 00.ctdb OK 0.011 Sat Dec 17 19:40:46 2016
436 01.reclock OK 0.010 Sat Dec 17 19:40:46 2016
437 05.system OK 0.030 Sat Dec 17 19:40:46 2016
438 06.nfs OK 0.014 Sat Dec 17 19:40:46 2016
439 10.interface OK 0.041 Sat Dec 17 19:40:46 2016
440 11.natgw OK 0.008 Sat Dec 17 19:40:46 2016
441 11.routing OK 0.007 Sat Dec 17 19:40:46 2016
442 13.per_ip_routing OK 0.007 Sat Dec 17 19:40:46 2016
443 20.multipathd OK 0.007 Sat Dec 17 19:40:46 2016
444 31.clamd OK 0.007 Sat Dec 17 19:40:46 2016
445 40.vsftpd OK 0.013 Sat Dec 17 19:40:46 2016
446 41.httpd OK 0.015 Sat Dec 17 19:40:46 2016
447 49.winbind OK 0.022 Sat Dec 17 19:40:46 2016
448 50.samba ERROR 0.077 Sat Dec 17 19:40:46 2016
449 OUTPUT: ERROR: samba tcp port 445 is not responding
450 listvars
453 List all tuneable variables, except the values of the obsolete tunables
454 like VacuumMinInterval. The obsolete tunables can be retrieved only
455 explicitly with the "ctdb getvar" command.
456 Example
458 # ctdb listvars
459 SeqnumInterval = 1000
460 ControlTimeout = 60
461 TraverseTimeout = 20
462 KeepaliveInterval = 5
463 KeepaliveLimit = 5
464 RecoverTimeout = 120
465 RecoverInterval = 1
466 ElectionTimeout = 3
467 TakeoverTimeout = 9
468 MonitorInterval = 15
469 TickleUpdateInterval = 20
470 EventScriptTimeout = 30
471 MonitorTimeoutCount = 20
472 RecoveryGracePeriod = 120
473 RecoveryBanPeriod = 300
474 DatabaseHashSize = 100001
475 DatabaseMaxDead = 5
476 RerecoveryTimeout = 10
477 EnableBans = 1
478 NoIPFailback = 0
479 VerboseMemoryNames = 0
480 RecdPingTimeout = 60
481 RecdFailCount = 10
482 LogLatencyMs = 0
483 RecLockLatencyMs = 1000
484 RecoveryDropAllIPs = 120
485 VacuumInterval = 10
486 VacuumMaxRunTime = 120
487 RepackLimit = 10000
488 VacuumLimit = 5000
489 VacuumFastPathCount = 60
490 MaxQueueDropMsg = 1000000
491 AllowUnhealthyDBRead = 0
492 StatHistoryInterval = 1
493 DeferredAttachTO = 120
494 AllowClientDBAttach = 1
495 RecoverPDBBySeqNum = 1
496 DeferredRebalanceOnNodeAdd = 300
497 FetchCollapse = 1
498 HopcountMakeSticky = 50
499 StickyDuration = 600
500 StickyPindown = 200
501 NoIPTakeover = 0
502 DBRecordCountWarn = 100000
503 DBRecordSizeWarn = 10000000
504 DBSizeWarn = 100000000
505 PullDBPreallocation = 10485760
506 LockProcessesPerDB = 200
507 RecBufferSizeLimit = 1000000
508 QueueBufferSize = 1024
509 IPAllocAlgorithm = 2
510 getvar NAME
513 Get the runtime value of a tuneable variable.
514 Example
516 # ctdb getvar MonitorInterval
517 MonitorInterval = 15
518 setvar NAME VALUE
521 Set the runtime value of a tuneable variable.
522 Example
524 # ctdb setvar MonitorInterval 20
525 lvs {master|list|status}
528 This command shows different aspects of LVS status. For an overview of
529 CTDB's LVS functionality please see the LVS section in ctdb(7).
530 master
532 Shows the PNN of the current LVS master node.
533 Example output:
535 2
537 list
540 Lists the currently usable LVS nodes.
541 Example output:
543 2 10.0.0.13
545 3 10.0.0.14
546 status
549 List the nodes in the current LVS group and their status.
550 Example output:
552 pnn:0 10.0.0.11 UNHEALTHY (THIS NODE)
554 pnn:1 10.0.0.12 UNHEALTHY
555 pnn:2 10.0.0.13 OK
556 pnn:3 10.0.0.14 OK
557 getcapabilities
560 This command shows the capabilities of the current node. See the
561 CAPABILITIES section in ctdb(7) for more details.
562 Example output:
564 RECMASTER: YES
566 LMASTER: YES
567 statistics
570 Collect statistics from the CTDB daemon about how many calls it has
571 served. Information about various fields in statistics can be found in
572 ctdb-statistics(7).
573 Example
575 # ctdb statistics
576 CTDB version 1
577 Current time of statistics : Tue Mar 8 15:18:51 2016
578 Statistics collected since : (003 21:31:32) Fri Mar 4 17:47:19 2016
579 num_clients 9
580 frozen 0
581 recovering 0
582 num_recoveries 2
583 client_packets_sent 8170534
584 client_packets_recv 7166132
585 node_packets_sent 16549998
586 node_packets_recv 5244418
587 keepalive_packets_sent 201969
588 keepalive_packets_recv 201969
589 node
590 req_call 26
591 reply_call 0
592 req_dmaster 9
593 reply_dmaster 12
594 reply_error 0
595 req_message 1339231
596 req_control 8177506
597 reply_control 6831284
598 client
599 req_call 15
600 req_message 334809
601 req_control 6831308
602 timeouts
603 call 0
604 control 0
605 traverse 0
606 locks
607 num_calls 8
608 num_current 0
609 num_pending 0
610 num_failed 0
611 total_calls 15
612 pending_calls 0
613 childwrite_calls 0
614 pending_childwrite_calls 0
615 memory_used 394879
616 max_hop_count 1
617 total_ro_delegations 0
618 total_ro_revokes 0
619 hop_count_buckets: 8 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0
620 lock_buckets: 0 0 8 0 0 0 0 0 0 0 0 0 0 0 0 0
621 locks_latency MIN/AVG/MAX 0.010005/0.010418/0.011010 sec out of 8
622 reclock_ctdbd MIN/AVG/MAX 0.002538/0.002538/0.002538 sec out of 1
623 reclock_recd MIN/AVG/MAX 0.000000/0.000000/0.000000 sec out of 0
624 call_latency MIN/AVG/MAX 0.000044/0.002142/0.011702 sec out of 15
625 childwrite_latency MIN/AVG/MAX 0.000000/0.000000/0.000000 sec out of 0
626 statisticsreset
629 This command is used to clear all statistics counters in a node.
630 Example: ctdb statisticsreset
632 dbstatistics DB
634 Display statistics about the database DB. Information about various
635 fields in dbstatistics can be found in ctdb-statistics(7).
636 Example
638 # ctdb dbstatistics locking.tdb
639 DB Statistics: locking.tdb
640 ro_delegations 0
641 ro_revokes 0
642 locks
643 total 14356
644 failed 0
645 current 0
646 pending 0
647 hop_count_buckets: 28087 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0
648 lock_buckets: 0 14188 38 76 32 19 3 0 0 0 0 0 0 0 0 0
649 locks_latency MIN/AVG/MAX 0.001066/0.012686/4.202292 sec out of 14356
650 vacuum_latency MIN/AVG/MAX 0.000472/0.002207/15.243570 sec out of 224530
651 Num Hot Keys: 1
652 Count:8 Key:ff5bd7cb3ee3822edc1f0000000000000000000000000000
653 getreclock
656 Show details of the recovery lock, if any.
657 Example output:
659 /clusterfs/.ctdb/recovery.lock
661 getdebug
664 Get the current debug level for the node. the debug level controls what
665 information is written to the log file.
666 The debug levels are mapped to the corresponding syslog levels. When a
668 debug level is set, only those messages at that level and higher levels
669 will be printed.
670 The list of debug levels from highest to lowest are :
672 ERROR WARNING NOTICE INFO DEBUG
674 setdebug DEBUGLEVEL
676 Set the debug level of a node. This controls what information will be
677 logged.
678 The debuglevel is one of ERROR WARNING NOTICE INFO DEBUG
680 getpid
682 This command will return the process id of the ctdb daemon.
683 disable
685 This command is used to administratively disable a node in the cluster.
686 A disabled node will still participate in the cluster and host
687 clustered TDB records but its public ip address has been taken over by
688 a different node and it no longer hosts any services.
689 enable
691 Re-enable a node that has been administratively disabled.
692 stop
694 This command is used to administratively STOP a node in the cluster. A
695 STOPPED node is connected to the cluster but will not host any public
696 ip addresse, nor does it participate in the VNNMAP. The difference
697 between a DISABLED node and a STOPPED node is that a STOPPED node does
698 not host any parts of the database which means that a recovery is
699 required to stop/continue nodes.
700 continue
702 Re-start a node that has been administratively stopped.
703 addip IPADDR/mask IFACE
705 This command is used to add a new public ip to a node during runtime.
706 It should be followed by a ctdb ipreallocate. This allows public
707 addresses to be added to a cluster without having to restart the ctdb
708 daemons.
709 Note that this only updates the runtime instance of ctdb. Any changes
711 will be lost next time ctdb is restarted and the public addresses file
712 is re-read. If you want this change to be permanent you must also
713 update the public addresses file manually.
714 delip IPADDR
716 This command flags IPADDR for deletion from a node at runtime. It
717 should be followed by a ctdb ipreallocate. If IPADDR is currently
718 hosted by the node it is being removed from, this ensures that the IP
719 will first be failed over to another node, if possible, and that it is
720 then actually removed.
721 Note that this only updates the runtime instance of CTDB. Any changes
723 will be lost next time CTDB is restarted and the public addresses file
724 is re-read. If you want this change to be permanent you must also
725 update the public addresses file manually.
726 moveip IPADDR PNN
728 This command can be used to manually fail a public ip address to a
729 specific node.
730 In order to manually override the "automatic" distribution of public ip
732 addresses that ctdb normally provides, this command only works when you
733 have changed the tunables for the daemon to:
734 IPAllocAlgorithm != 0
736 NoIPFailback = 1
738 shutdown
740 This command will shutdown a specific CTDB daemon.
741 setlmasterrole on|off
743 This command is used ot enable/disable the LMASTER capability for a
744 node at runtime. This capability determines whether or not a node can
745 be used as an LMASTER for records in the database. A node that does not
746 have the LMASTER capability will not show up in the vnnmap.
747 Nodes will by default have this capability, but it can be stripped off
749 nodes by the setting in the sysconfig file or by using this command.
750 Once this setting has been enabled/disabled, you need to perform a
752 recovery for it to take effect.
753 See also "ctdb getcapabilities"
755 setrecmasterrole on|off
757 This command is used ot enable/disable the RECMASTER capability for a
758 node at runtime. This capability determines whether or not a node can
759 be used as an RECMASTER for the cluster. A node that does not have the
760 RECMASTER capability can not win a recmaster election. A node that
761 already is the recmaster for the cluster when the capability is
762 stripped off the node will remain the recmaster until the next cluster
763 election.
764 Nodes will by default have this capability, but it can be stripped off
766 nodes by the setting in the sysconfig file or by using this command.
767 See also "ctdb getcapabilities"
769 reloadnodes
771 This command is used when adding new nodes, or removing existing nodes
772 from an existing cluster.
773 Procedure to add nodes:
775 1. To expand an existing cluster, first ensure with ctdb status that
777 all nodes are up and running and that they are all healthy. Do not
778 try to expand a cluster unless it is completely healthy!
779 2. On all nodes, edit /etc/ctdb/nodes and add the new nodes at the end
781 of this file.
782 3. Verify that all the nodes have identical /etc/ctdb/nodes files
784 after adding the new nodes.
785 4. Run ctdb reloadnodes to force all nodes to reload the nodes file.
787 5. Use ctdb status on all nodes and verify that they now show the
789 additional nodes.
790 6. Install and configure the new node and bring it online.
792 Procedure to remove nodes:
794 1. To remove nodes from an existing cluster, first ensure with ctdb
796 status that all nodes, except the node to be deleted, are up and
797 running and that they are all healthy. Do not try to remove nodes
798 from a cluster unless the cluster is completely healthy!
799 2. Shutdown and power off the node to be removed.
801 3. On all other nodes, edit the /etc/ctdb/nodes file and comment out
803 the nodes to be removed. Do not delete the lines for the deleted
804 nodes, just comment them out by adding a '#' at the beginning of
805 the lines.
806 4. Run ctdb reloadnodes to force all nodes to reload the nodes file.
808 5. Use ctdb status on all nodes and verify that the deleted nodes are
810 no longer listed.
811 reloadips [PNN-LIST]
813 This command reloads the public addresses configuration file on the
814 specified nodes. When it completes addresses will be reconfigured and
815 reassigned across the cluster as necessary.
816 This command is currently unable to make changes to the netmask or
818 interfaces associated with existing addresses. Such changes must be
819 made in 2 steps by deleting addresses in question and re-adding then.
820 Unfortunately this will disrupt connections to the changed addresses.
821 getdbmap
823 This command lists all clustered TDB databases that the CTDB daemon has
824 attached to. Some databases are flagged as PERSISTENT, this means that
825 the database stores data persistently and the data will remain across
826 reboots. One example of such a database is secrets.tdb where
827 information about how the cluster was joined to the domain is stored.
828 Some database are flagged as REPLICATED, this means that the data in
829 that database is replicated across all the nodes. But the data will not
830 remain across reboots. This type of database is used by CTDB to store
831 it's internal state.
832 If a PERSISTENT database is not in a healthy state the database is
834 flagged as UNHEALTHY. If there's at least one completely healthy node
835 running in the cluster, it's possible that the content is restored by a
836 recovery run automaticly. Otherwise an administrator needs to analyze
837 the problem.
838 See also "ctdb getdbstatus", "ctdb backupdb", "ctdb restoredb", "ctdb
840 dumpbackup", "ctdb wipedb", "ctdb setvar AllowUnhealthyDBRead 1" and
841 (if samba or tdb-utils are installed) "tdbtool check".
842 Most databases are not persistent and only store the state information
844 that the currently running samba daemons need. These databases are
845 always wiped when ctdb/samba starts and when a node is rebooted.
846 Example
848 # ctdb getdbmap
849 Number of databases:10
850 dbid:0x435d3410 name:notify.tdb path:/var/lib/ctdb/notify.tdb.0
851 dbid:0x42fe72c5 name:locking.tdb path:/var/lib/ctdb/locking.tdb.0
852 dbid:0x1421fb78 name:brlock.tdb path:/var/lib/ctdb/brlock.tdb.0
853 dbid:0x17055d90 name:connections.tdb path:/var/lib/ctdb/connections.tdb.0
854 dbid:0xc0bdde6a name:sessionid.tdb path:/var/lib/ctdb/sessionid.tdb.0
855 dbid:0x122224da name:test.tdb path:/var/lib/ctdb/test.tdb.0
856 dbid:0x2672a57f name:idmap2.tdb path:/var/lib/ctdb/persistent/idmap2.tdb.0 PERSISTENT
857 dbid:0xb775fff6 name:secrets.tdb path:/var/lib/ctdb/persistent/secrets.tdb.0 PERSISTENT
858 dbid:0xe98e08b6 name:group_mapping.tdb path:/var/lib/ctdb/persistent/group_mapping.tdb.0 PERSISTENT
859 dbid:0x7bbbd26c name:passdb.tdb path:/var/lib/ctdb/persistent/passdb.tdb.0 PERSISTENT
860 # ctdb getdbmap # example for unhealthy database
862 Number of databases:1
863 dbid:0xb775fff6 name:secrets.tdb path:/var/lib/ctdb/persistent/secrets.tdb.0 PERSISTENT UNHEALTHY
864 # ctdb -X getdbmap
866 |ID|Name|Path|Persistent|Unhealthy|
867 |0x7bbbd26c|passdb.tdb|/var/lib/ctdb/persistent/passdb.tdb.0|1|0|
868 backupdb DB FILE
871 Copy the contents of database DB to FILE. FILE can later be read back
872 using restoredb. This is mainly useful for backing up persistent
873 databases such as secrets.tdb and similar.
874 restoredb FILE [DB]
876 This command restores a persistent database that was previously backed
877 up using backupdb. By default the data will be restored back into the
878 same database as it was created from. By specifying dbname you can
879 restore the data into a different database.
880 setdbreadonly DB
882 This command will enable the read-only record support for a database.
883 This is an experimental feature to improve performance for contended
884 records primarily in locking.tdb and brlock.tdb. When enabling this
885 feature you must set it on all nodes in the cluster.
886 setdbsticky DB
888 This command will enable the sticky record support for the specified
889 database. This is an experimental feature to improve performance for
890 contended records primarily in locking.tdb and brlock.tdb. When
891 enabling this feature you must set it on all nodes in the cluster.
892
894 Internal commands are used by CTDB's scripts and are not required for
895 managing a CTDB cluster. Their parameters and behaviour are subject to
896 change.
897 gettickles IPADDR
899 Show TCP connections that are registered with CTDB to be "tickled" if
900 there is a failover.
901 gratarp IPADDR INTERFACE
903 Send out a gratuitous ARP for the specified interface through the
904 specified interface. This command is mainly used by the ctdb
905 eventscripts.
906 pdelete DB KEY
908 Delete KEY from DB.
909 pfetch DB KEY
911 Print the value associated with KEY in DB.
912 pstore DB KEY FILE
914 Store KEY in DB with contents of FILE as the associated value.
915 ptrans DB [FILE]
917 Read a list of key-value pairs, one per line from FILE, and store them
918 in DB using a single transaction. An empty value is equivalent to
919 deleting the given key.
920 The key and value should be separated by spaces or tabs. Each key/value
922 should be a printable string enclosed in double-quotes.
923 runstate [setup|first_recovery|startup|running]
925 Print the runstate of the specified node. Runstates are used to
926 serialise important state transitions in CTDB, particularly during
927 startup.
928 If one or more optional runstate arguments are specified then the node
930 must be in one of these runstates for the command to succeed.
931 Example
933 # ctdb runstate
934 RUNNING
935 setifacelink IFACE up|down
938 Set the internal state of network interface IFACE. This is typically
939 used in the 10.interface script in the "monitor" event.
940 Example: ctdb setifacelink eth0 up
942 tickle
944 Read a list of TCP connections, one per line, from standard input and
945 send a TCP tickle to the source host for each connection. A connection
946 is specified as:
947 SRC-IPADDR:SRC-PORT DST-IPADDR:DST-PORT
949 A single connection can be specified on the command-line rather than on
952 standard input.
953 A TCP tickle is a TCP ACK packet with an invalid sequence and
955 acknowledge number and will when received by the source host result in
956 it sending an immediate correct ACK back to the other end.
957 TCP tickles are useful to "tickle" clients after a IP failover has
959 occurred since this will make the client immediately recognize the TCP
960 connection has been disrupted and that the client will need to
961 reestablish. This greatly speeds up the time it takes for a client to
962 detect and reestablish after an IP failover in the ctdb cluster.
963 version
965 Display the CTDB version.
966
968 These commands are primarily used for CTDB development and testing and
969 should not be used for normal administration.
970 OPTIONS
972 --print-emptyrecords
973 This enables printing of empty records when dumping databases with
974 the catdb, cattbd and dumpdbbackup commands. Records with empty
975 data segment are considered deleted by ctdb and cleaned by the
976 vacuuming mechanism, so this switch can come in handy for debugging
977 the vacuuming behaviour.
978 --print-datasize
980 This lets database dumps (catdb, cattdb, dumpdbbackup) print the
981 size of the record data instead of dumping the data contents.
982 --print-lmaster
984 This lets catdb print the lmaster for each record.
985 --print-hash
987 This lets database dumps (catdb, cattdb, dumpdbbackup) print the
988 hash for each record.
989 --print-recordflags
991 This lets catdb and dumpdbbackup print the record flags for each
992 record. Note that cattdb always prints the flags.
993 process-exists PID [SRVID]
995 This command checks if a specific process exists on the CTDB host. This
996 is mainly used by Samba to check if remote instances of samba are still
997 running or not. When the optional SRVID argument is specified, the
998 command check if a specific process exists on the CTDB host and has
999 registered for specified SRVID.
1000 getdbstatus DB
1002 This command displays more details about a database.
1003 Example
1005 # ctdb getdbstatus test.tdb.0
1006 dbid: 0x122224da
1007 name: test.tdb
1008 path: /var/lib/ctdb/test.tdb.0
1009 PERSISTENT: no
1010 HEALTH: OK
1011 # ctdb getdbstatus registry.tdb # with a corrupted TDB
1013 dbid: 0xf2a58948
1014 name: registry.tdb
1015 path: /var/lib/ctdb/persistent/registry.tdb.0
1016 PERSISTENT: yes
1017 HEALTH: NO-HEALTHY-NODES - ERROR - Backup of corrupted TDB in '/var/lib/ctdb/persistent/registry.tdb.0.corrupted.20091208091949.0Z'
1018 catdb DB
1021 Print a dump of the clustered TDB database DB.
1022 cattdb DB
1024 Print a dump of the contents of the local TDB database DB.
1025 dumpdbbackup FILE
1027 Print a dump of the contents from database backup FILE, similar to
1028 catdb.
1029 wipedb DB
1031 Remove all contents of database DB.
1032 recover
1034 This command will trigger the recovery daemon to do a cluster recovery.
1035 ipreallocate, sync
1037 This command will force the recovery master to perform a full ip
1038 reallocation process and redistribute all ip addresses. This is useful
1039 to "reset" the allocations back to its default state if they have been
1040 changed using the "moveip" command. While a "recover" will also perform
1041 this reallocation, a recovery is much more hevyweight since it will
1042 also rebuild all the databases.
1043 attach DBNAME [persistent|replicated]
1045 Create a new CTDB database called DBNAME and attach to it on all nodes.
1046 detach DB-LIST
1048 Detach specified non-persistent database(s) from the cluster. This
1049 command will disconnect specified database(s) on all nodes in the
1050 cluster. This command should only be used when none of the specified
1051 database(s) are in use.
1052 All nodes should be active and tunable AllowClientDBAccess should be
1054 disabled on all nodes before detaching databases.
1055 dumpmemory
1057 This is a debugging command. This command will make the ctdb daemon to
1058 write a fill memory allocation map to standard output.
1059 rddumpmemory
1061 This is a debugging command. This command will dump the talloc memory
1062 allocation tree for the recovery daemon to standard output.
1063 ban BANTIME
1065 Administratively ban a node for BANTIME seconds. The node will be
1066 unbanned after BANTIME seconds have elapsed.
1067 A banned node does not participate in the cluster. It does not host any
1069 records for the clustered TDB and does not host any public IP
1070 addresses.
1071 Nodes are automatically banned if they misbehave. For example, a node
1073 may be banned if it causes too many cluster recoveries.
1074 To administratively exclude a node from a cluster use the stop command.
1076 unban
1078 This command is used to unban a node that has either been
1079 administratively banned using the ban command or has been automatically
1080 banned.
1081
1083 ctdbd(1), onnode(1), ctdb(7), ctdb-statistics(7), ctdb-tunables(7),
1084 http://ctdb.samba.org/
1085
1087 This documentation was written by Ronnie Sahlberg, Amitay Isaacs,
1088 Martin Schwenke
1089
1091 Copyright © 2007 Andrew Tridgell, Ronnie Sahlberg
1092 This program is free software; you can redistribute it and/or modify it
1094 under the terms of the GNU General Public License as published by the
1095 Free Software Foundation; either version 3 of the License, or (at your
1096 option) any later version.
1097 This program is distributed in the hope that it will be useful, but
1099 WITHOUT ANY WARRANTY; without even the implied warranty of
1100 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
1101 General Public License for more details.
1102 You should have received a copy of the GNU General Public License along
1104 with this program; if not, see http://www.gnu.org/licenses.
1105ctdb 12/16/2019 CTDB(1)