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 usage 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 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 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 VacuumFastPathCount = 60
489 MaxQueueDropMsg = 1000000
490 AllowUnhealthyDBRead = 0
491 StatHistoryInterval = 1
492 DeferredAttachTO = 120
493 AllowClientDBAttach = 1
494 RecoverPDBBySeqNum = 1
495 DeferredRebalanceOnNodeAdd = 300
496 FetchCollapse = 1
497 HopcountMakeSticky = 50
498 StickyDuration = 600
499 StickyPindown = 200
500 NoIPTakeover = 0
501 DBRecordCountWarn = 100000
502 DBRecordSizeWarn = 10000000
503 DBSizeWarn = 100000000
504 PullDBPreallocation = 10485760
505 LockProcessesPerDB = 200
506 RecBufferSizeLimit = 1000000
507 QueueBufferSize = 1024
508 IPAllocAlgorithm = 2
509 getvar NAME
512 Get the runtime value of a tuneable variable.
513 Example
515 # ctdb getvar MonitorInterval
516 MonitorInterval = 15
517 setvar NAME VALUE
520 Set the runtime value of a tuneable variable.
521 Example
523 # ctdb setvar MonitorInterval 20
524 lvs {master|list|status}
527 This command shows different aspects of LVS status. For an overview of
528 CTDB's LVS functionality please see the LVS section in ctdb(7).
529 master
531 Shows the PNN of the current LVS master node.
532 Example output:
534 2
536 list
539 Lists the currently usable LVS nodes.
540 Example output:
542 2 10.0.0.13
544 3 10.0.0.14
545 status
548 List the nodes in the current LVS group and their status.
549 Example output:
551 pnn:0 10.0.0.11 UNHEALTHY (THIS NODE)
553 pnn:1 10.0.0.12 UNHEALTHY
554 pnn:2 10.0.0.13 OK
555 pnn:3 10.0.0.14 OK
556 getcapabilities
559 This command shows the capabilities of the current node. See the
560 CAPABILITIES section in ctdb(7) for more details.
561 Example output:
563 RECMASTER: YES
565 LMASTER: YES
566 statistics
569 Collect statistics from the CTDB daemon about how many calls it has
570 served. Information about various fields in statistics can be found in
571 ctdb-statistics(7).
572 Example
574 # ctdb statistics
575 CTDB version 1
576 Current time of statistics : Tue Mar 8 15:18:51 2016
577 Statistics collected since : (003 21:31:32) Fri Mar 4 17:47:19 2016
578 num_clients 9
579 frozen 0
580 recovering 0
581 num_recoveries 2
582 client_packets_sent 8170534
583 client_packets_recv 7166132
584 node_packets_sent 16549998
585 node_packets_recv 5244418
586 keepalive_packets_sent 201969
587 keepalive_packets_recv 201969
588 node
589 req_call 26
590 reply_call 0
591 req_dmaster 9
592 reply_dmaster 12
593 reply_error 0
594 req_message 1339231
595 req_control 8177506
596 reply_control 6831284
597 client
598 req_call 15
599 req_message 334809
600 req_control 6831308
601 timeouts
602 call 0
603 control 0
604 traverse 0
605 locks
606 num_calls 8
607 num_current 0
608 num_pending 0
609 num_failed 0
610 total_calls 15
611 pending_calls 0
612 childwrite_calls 0
613 pending_childwrite_calls 0
614 memory_used 394879
615 max_hop_count 1
616 total_ro_delegations 0
617 total_ro_revokes 0
618 hop_count_buckets: 8 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0
619 lock_buckets: 0 0 8 0 0 0 0 0 0 0 0 0 0 0 0 0
620 locks_latency MIN/AVG/MAX 0.010005/0.010418/0.011010 sec out of 8
621 reclock_ctdbd MIN/AVG/MAX 0.002538/0.002538/0.002538 sec out of 1
622 reclock_recd MIN/AVG/MAX 0.000000/0.000000/0.000000 sec out of 0
623 call_latency MIN/AVG/MAX 0.000044/0.002142/0.011702 sec out of 15
624 childwrite_latency MIN/AVG/MAX 0.000000/0.000000/0.000000 sec out of 0
625 statisticsreset
628 This command is used to clear all statistics counters in a node.
629 Example: ctdb statisticsreset
631 dbstatistics DB
633 Display statistics about the database DB. Information about various
634 fields in dbstatistics can be found in ctdb-statistics(7).
635 Example
637 # ctdb dbstatistics locking.tdb
638 DB Statistics: locking.tdb
639 ro_delegations 0
640 ro_revokes 0
641 locks
642 total 14356
643 failed 0
644 current 0
645 pending 0
646 hop_count_buckets: 28087 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0
647 lock_buckets: 0 14188 38 76 32 19 3 0 0 0 0 0 0 0 0 0
648 locks_latency MIN/AVG/MAX 0.001066/0.012686/4.202292 sec out of 14356
649 vacuum_latency MIN/AVG/MAX 0.000472/0.002207/15.243570 sec out of 224530
650 Num Hot Keys: 1
651 Count:8 Key:ff5bd7cb3ee3822edc1f0000000000000000000000000000
652 getreclock
655 Show details of the recovery lock, if any.
656 Example output:
658 /clusterfs/.ctdb/recovery.lock
660 getdebug
663 Get the current debug level for the node. the debug level controls what
664 information is written to the log file.
665 The debug levels are mapped to the corresponding syslog levels. When a
667 debug level is set, only those messages at that level and higher levels
668 will be printed.
669 The list of debug levels from highest to lowest are :
671 ERROR WARNING NOTICE INFO DEBUG
673 setdebug DEBUGLEVEL
675 Set the debug level of a node. This controls what information will be
676 logged.
677 The debuglevel is one of ERROR WARNING NOTICE INFO DEBUG
679 getpid
681 This command will return the process id of the ctdb daemon.
682 disable
684 This command is used to administratively disable a node in the cluster.
685 A disabled node will still participate in the cluster and host
686 clustered TDB records but its public ip address has been taken over by
687 a different node and it no longer hosts any services.
688 enable
690 Re-enable a node that has been administratively disabled.
691 stop
693 This command is used to administratively STOP a node in the cluster. A
694 STOPPED node is connected to the cluster but will not host any public
695 ip addresse, nor does it participate in the VNNMAP. The difference
696 between a DISABLED node and a STOPPED node is that a STOPPED node does
697 not host any parts of the database which means that a recovery is
698 required to stop/continue nodes.
699 continue
701 Re-start a node that has been administratively stopped.
702 addip IPADDR/mask IFACE
704 This command is used to add a new public ip to a node during runtime.
705 It should be followed by a ctdb ipreallocate. This allows public
706 addresses to be added to a cluster without having to restart the ctdb
707 daemons.
708 Note that this only updates the runtime instance of ctdb. Any changes
710 will be lost next time ctdb is restarted and the public addresses file
711 is re-read. If you want this change to be permanent you must also
712 update the public addresses file manually.
713 delip IPADDR
715 This command flags IPADDR for deletion from a node at runtime. It
716 should be followed by a ctdb ipreallocate. If IPADDR is currently
717 hosted by the node it is being removed from, this ensures that the IP
718 will first be failed over to another node, if possible, and that it is
719 then actually removed.
720 Note that this only updates the runtime instance of CTDB. Any changes
722 will be lost next time CTDB is restarted and the public addresses file
723 is re-read. If you want this change to be permanent you must also
724 update the public addresses file manually.
725 moveip IPADDR PNN
727 This command can be used to manually fail a public ip address to a
728 specific node.
729 In order to manually override the "automatic" distribution of public ip
731 addresses that ctdb normally provides, this command only works when you
732 have changed the tunables for the daemon to:
733 IPAllocAlgorithm != 0
735 NoIPFailback = 1
737 shutdown
739 This command will shutdown a specific CTDB daemon.
740 setlmasterrole on|off
742 This command is used to enable/disable the LMASTER capability for a
743 node at runtime. This capability determines whether or not a node can
744 be used as an LMASTER for records in the database. A node that does not
745 have the LMASTER capability will not show up in the vnnmap.
746 Nodes will by default have this capability, but it can be stripped off
748 nodes by the setting in the sysconfig file or by using this command.
749 Once this setting has been enabled/disabled, you need to perform a
751 recovery for it to take effect.
752 See also "ctdb getcapabilities"
754 setrecmasterrole on|off
756 This command is used to enable/disable the RECMASTER capability for a
757 node at runtime. This capability determines whether or not a node can
758 be used as an RECMASTER for the cluster. A node that does not have the
759 RECMASTER capability can not win a recmaster election. A node that
760 already is the recmaster for the cluster when the capability is
761 stripped off the node will remain the recmaster until the next cluster
762 election.
763 Nodes will by default have this capability, but it can be stripped off
765 nodes by the setting in the sysconfig file or by using this command.
766 See also "ctdb getcapabilities"
768 reloadnodes
770 This command is used when adding new nodes, or removing existing nodes
771 from an existing cluster.
772 Procedure to add nodes:
774 1. To expand an existing cluster, first ensure with ctdb status that
776 all nodes are up and running and that they are all healthy. Do not
777 try to expand a cluster unless it is completely healthy!
778 2. On all nodes, edit /etc/ctdb/nodes and add the new nodes at the end
780 of this file.
781 3. Verify that all the nodes have identical /etc/ctdb/nodes files
783 after adding the new nodes.
784 4. Run ctdb reloadnodes to force all nodes to reload the nodes file.
786 5. Use ctdb status on all nodes and verify that they now show the
788 additional nodes.
789 6. Install and configure the new node and bring it online.
791 Procedure to remove nodes:
793 1. To remove nodes from an existing cluster, first ensure with ctdb
795 status that all nodes, except the node to be deleted, are up and
796 running and that they are all healthy. Do not try to remove nodes
797 from a cluster unless the cluster is completely healthy!
798 2. Shutdown and power off the node to be removed.
800 3. On all other nodes, edit the /etc/ctdb/nodes file and comment out
802 the nodes to be removed. Do not delete the lines for the deleted
803 nodes, just comment them out by adding a '#' at the beginning of
804 the lines.
805 4. Run ctdb reloadnodes to force all nodes to reload the nodes file.
807 5. Use ctdb status on all nodes and verify that the deleted nodes are
809 no longer listed.
810 reloadips [PNN-LIST]
812 This command reloads the public addresses configuration file on the
813 specified nodes. When it completes addresses will be reconfigured and
814 reassigned across the cluster as necessary.
815 This command is currently unable to make changes to the netmask or
817 interfaces associated with existing addresses. Such changes must be
818 made in 2 steps by deleting addresses in question and re-adding then.
819 Unfortunately this will disrupt connections to the changed addresses.
820 getdbmap
822 This command lists all clustered TDB databases that the CTDB daemon has
823 attached to. Some databases are flagged as PERSISTENT, this means that
824 the database stores data persistently and the data will remain across
825 reboots. One example of such a database is secrets.tdb where
826 information about how the cluster was joined to the domain is stored.
827 Some database are flagged as REPLICATED, this means that the data in
828 that database is replicated across all the nodes. But the data will not
829 remain across reboots. This type of database is used by CTDB to store
830 it's internal state.
831 If a PERSISTENT database is not in a healthy state the database is
833 flagged as UNHEALTHY. If there's at least one completely healthy node
834 running in the cluster, it's possible that the content is restored by a
835 recovery run automatically. Otherwise an administrator needs to analyze
836 the problem.
837 See also "ctdb getdbstatus", "ctdb backupdb", "ctdb restoredb", "ctdb
839 dumpbackup", "ctdb wipedb", "ctdb setvar AllowUnhealthyDBRead 1" and
840 (if samba or tdb-utils are installed) "tdbtool check".
841 Most databases are not persistent and only store the state information
843 that the currently running samba daemons need. These databases are
844 always wiped when ctdb/samba starts and when a node is rebooted.
845 Example
847 # ctdb getdbmap
848 Number of databases:10
849 dbid:0x435d3410 name:notify.tdb path:/var/lib/ctdb/notify.tdb.0
850 dbid:0x42fe72c5 name:locking.tdb path:/var/lib/ctdb/locking.tdb.0
851 dbid:0x1421fb78 name:brlock.tdb path:/var/lib/ctdb/brlock.tdb.0
852 dbid:0x17055d90 name:connections.tdb path:/var/lib/ctdb/connections.tdb.0
853 dbid:0xc0bdde6a name:sessionid.tdb path:/var/lib/ctdb/sessionid.tdb.0
854 dbid:0x122224da name:test.tdb path:/var/lib/ctdb/test.tdb.0
855 dbid:0x2672a57f name:idmap2.tdb path:/var/lib/ctdb/persistent/idmap2.tdb.0 PERSISTENT
856 dbid:0xb775fff6 name:secrets.tdb path:/var/lib/ctdb/persistent/secrets.tdb.0 PERSISTENT
857 dbid:0xe98e08b6 name:group_mapping.tdb path:/var/lib/ctdb/persistent/group_mapping.tdb.0 PERSISTENT
858 dbid:0x7bbbd26c name:passdb.tdb path:/var/lib/ctdb/persistent/passdb.tdb.0 PERSISTENT
859 # ctdb getdbmap # example for unhealthy database
861 Number of databases:1
862 dbid:0xb775fff6 name:secrets.tdb path:/var/lib/ctdb/persistent/secrets.tdb.0 PERSISTENT UNHEALTHY
863 # ctdb -X getdbmap
865 |ID|Name|Path|Persistent|Unhealthy|
866 |0x7bbbd26c|passdb.tdb|/var/lib/ctdb/persistent/passdb.tdb.0|1|0|
867 backupdb DB FILE
870 Copy the contents of database DB to FILE. FILE can later be read back
871 using restoredb. This is mainly useful for backing up persistent
872 databases such as secrets.tdb and similar.
873 restoredb FILE [DB]
875 This command restores a persistent database that was previously backed
876 up using backupdb. By default the data will be restored back into the
877 same database as it was created from. By specifying dbname you can
878 restore the data into a different database.
879 setdbreadonly DB
881 This command will enable the read-only record support for a database.
882 This is an experimental feature to improve performance for contended
883 records primarily in locking.tdb and brlock.tdb. When enabling this
884 feature you must set it on all nodes in the cluster.
885 setdbsticky DB
887 This command will enable the sticky record support for the specified
888 database. This is an experimental feature to improve performance for
889 contended records primarily in locking.tdb and brlock.tdb. When
890 enabling this feature you must set it on all nodes in the cluster.
891
893 Internal commands are used by CTDB's scripts and are not required for
894 managing a CTDB cluster. Their parameters and behaviour are subject to
895 change.
896 gettickles IPADDR
898 Show TCP connections that are registered with CTDB to be "tickled" if
899 there is a failover.
900 gratarp IPADDR INTERFACE
902 Send out a gratuitous ARP for the specified interface through the
903 specified interface. This command is mainly used by the ctdb
904 eventscripts.
905 pdelete DB KEY
907 Delete KEY from DB.
908 pfetch DB KEY
910 Print the value associated with KEY in DB.
911 pstore DB KEY FILE
913 Store KEY in DB with contents of FILE as the associated value.
914 ptrans DB [FILE]
916 Read a list of key-value pairs, one per line from FILE, and store them
917 in DB using a single transaction. An empty value is equivalent to
918 deleting the given key.
919 The key and value should be separated by spaces or tabs. Each key/value
921 should be a printable string enclosed in double-quotes.
922 runstate [setup|first_recovery|startup|running]
924 Print the runstate of the specified node. Runstates are used to
925 serialise important state transitions in CTDB, particularly during
926 startup.
927 If one or more optional runstate arguments are specified then the node
929 must be in one of these runstates for the command to succeed.
930 Example
932 # ctdb runstate
933 RUNNING
934 setifacelink IFACE up|down
937 Set the internal state of network interface IFACE. This is typically
938 used in the 10.interface script in the "monitor" event.
939 Example: ctdb setifacelink eth0 up
941 tickle
943 Read a list of TCP connections, one per line, from standard input and
944 send a TCP tickle to the source host for each connection. A connection
945 is specified as:
946 SRC-IPADDR:SRC-PORT DST-IPADDR:DST-PORT
948 A single connection can be specified on the command-line rather than on
951 standard input.
952 A TCP tickle is a TCP ACK packet with an invalid sequence and
954 acknowledge number and will when received by the source host result in
955 it sending an immediate correct ACK back to the other end.
956 TCP tickles are useful to "tickle" clients after a IP failover has
958 occurred since this will make the client immediately recognize the TCP
959 connection has been disrupted and that the client will need to
960 reestablish. This greatly speeds up the time it takes for a client to
961 detect and reestablish after an IP failover in the ctdb cluster.
962 version
964 Display the CTDB version.
965
967 These commands are primarily used for CTDB development and testing and
968 should not be used for normal administration.
969 OPTIONS
971 --print-emptyrecords
972 This enables printing of empty records when dumping databases with
973 the catdb, cattbd and dumpdbbackup commands. Records with empty
974 data segment are considered deleted by ctdb and cleaned by the
975 vacuuming mechanism, so this switch can come in handy for debugging
976 the vacuuming behaviour.
977 --print-datasize
979 This lets database dumps (catdb, cattdb, dumpdbbackup) print the
980 size of the record data instead of dumping the data contents.
981 --print-lmaster
983 This lets catdb print the lmaster for each record.
984 --print-hash
986 This lets database dumps (catdb, cattdb, dumpdbbackup) print the
987 hash for each record.
988 --print-recordflags
990 This lets catdb and dumpdbbackup print the record flags for each
991 record. Note that cattdb always prints the flags.
992 process-exists PID [SRVID]
994 This command checks if a specific process exists on the CTDB host. This
995 is mainly used by Samba to check if remote instances of samba are still
996 running or not. When the optional SRVID argument is specified, the
997 command check if a specific process exists on the CTDB host and has
998 registered for specified SRVID.
999 getdbstatus DB
1001 This command displays more details about a database.
1002 Example
1004 # ctdb getdbstatus test.tdb.0
1005 dbid: 0x122224da
1006 name: test.tdb
1007 path: /var/lib/ctdb/test.tdb.0
1008 PERSISTENT: no
1009 HEALTH: OK
1010 # ctdb getdbstatus registry.tdb # with a corrupted TDB
1012 dbid: 0xf2a58948
1013 name: registry.tdb
1014 path: /var/lib/ctdb/persistent/registry.tdb.0
1015 PERSISTENT: yes
1016 HEALTH: NO-HEALTHY-NODES - ERROR - Backup of corrupted TDB in '/var/lib/ctdb/persistent/registry.tdb.0.corrupted.20091208091949.0Z'
1017 catdb DB
1020 Print a dump of the clustered TDB database DB.
1021 cattdb DB
1023 Print a dump of the contents of the local TDB database DB.
1024 dumpdbbackup FILE
1026 Print a dump of the contents from database backup FILE, similar to
1027 catdb.
1028 wipedb DB
1030 Remove all contents of database DB.
1031 recover
1033 This command will trigger the recovery daemon to do a cluster recovery.
1034 ipreallocate, sync
1036 This command will force the recovery master to perform a full ip
1037 reallocation process and redistribute all ip addresses. This is useful
1038 to "reset" the allocations back to its default state if they have been
1039 changed using the "moveip" command. While a "recover" will also perform
1040 this reallocation, a recovery is much more hevyweight since it will
1041 also rebuild all the databases.
1042 attach DBNAME [persistent|replicated]
1044 Create a new CTDB database called DBNAME and attach to it on all nodes.
1045 detach DB-LIST
1047 Detach specified non-persistent database(s) from the cluster. This
1048 command will disconnect specified database(s) on all nodes in the
1049 cluster. This command should only be used when none of the specified
1050 database(s) are in use.
1051 All nodes should be active and tunable AllowClientDBAccess should be
1053 disabled on all nodes before detaching databases.
1054 dumpmemory
1056 This is a debugging command. This command will make the ctdb daemon to
1057 write a fill memory allocation map to standard output.
1058 rddumpmemory
1060 This is a debugging command. This command will dump the talloc memory
1061 allocation tree for the recovery daemon to standard output.
1062 ban BANTIME
1064 Administratively ban a node for BANTIME seconds. The node will be
1065 unbanned after BANTIME seconds have elapsed.
1066 A banned node does not participate in the cluster. It does not host any
1068 records for the clustered TDB and does not host any public IP
1069 addresses.
1070 Nodes are automatically banned if they misbehave. For example, a node
1072 may be banned if it causes too many cluster recoveries.
1073 To administratively exclude a node from a cluster use the stop command.
1075 unban
1077 This command is used to unban a node that has either been
1078 administratively banned using the ban command or has been automatically
1079 banned.
1080
1082 ctdbd(1), onnode(1), ctdb(7), ctdb-statistics(7), ctdb-tunables(7),
1083 http://ctdb.samba.org/
1084
1086 This documentation was written by Ronnie Sahlberg, Amitay Isaacs,
1087 Martin Schwenke
1088
1090 Copyright © 2007 Andrew Tridgell, Ronnie Sahlberg
1091 This program is free software; you can redistribute it and/or modify it
1093 under the terms of the GNU General Public License as published by the
1094 Free Software Foundation; either version 3 of the License, or (at your
1095 option) any later version.
1096 This program is distributed in the hope that it will be useful, but
1098 WITHOUT ANY WARRANTY; without even the implied warranty of
1099 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
1100 General Public License for more details.
1101 You should have received a copy of the GNU General Public License along
1103 with this program; if not, see http://www.gnu.org/licenses.
1104ctdb 04/28/2020 CTDB(1)