1XTRABACKUP(1) Percona XtraBackup XTRABACKUP(1)
2
6 xtrabackup - Percona XtraBackup 2.3 Documentation
7 The xtrabackup binary is a compiled C program that is linked with the
9 InnoDB libraries and the standard MySQL client libraries. The InnoDB
10 libraries provide functionality necessary to apply a log to data files,
11 and the MySQL client libraries provide command-line option parsing,
12 configuration file parsing, and so on to give the binary a familiar
13 look and feel.
14 The tool runs in either --backup or --prepare mode, corresponding to
16 the two main functions it performs. There are several variations on
17 these functions to accomplish different tasks, and there are two less
18 commonly used modes, --stats and --print-param.
19
21 Creating a Backup
22 To create a backup, run xtrabackup with the --backup option. You also
23 need to specify a --target_dir option, which is where the backup will
24 be stored, and a --datadir option, which is where the MySQL data is
25 stored. If the InnoDB data or log files aren't stored in the same
26 directory, you might need to specify the location of those, too. If the
27 target directory does not exist, xtrabackup creates it. If the direc‐
28 tory does exist and is empty, xtrabackup will succeed. xtrabackup will
29 not overwrite existing files, it will fail with operating system error
30 17, file exists.
31 The tool changes its working directory to the data directory and per‐
33 forms two primary tasks to complete the backup:
34 · It starts a log-copying thread in the background. This thread watches
36 the InnoDB log files, and when they change, it copies the changed
37 blocks to a file called xtrabackup_logfile in the backup target
38 directory. This is necessary because the backup might take a long
39 time, and the recovery process needs all of the log file entries from
40 the beginning to the end of the backup.
41 · It copies the InnoDB data files to the target directory. This is not
43 a simple file copy; it opens and reads the files similarly to the way
44 InnoDB does, by reading the data dictionary and copying them a page
45 at a time.
46 When the data files are finished copying, xtrabackup stops the
48 log-copying thread, and creates a files in the target directory called
49 xtrabackup_checkpoints, which contains the type of backup performed,
50 the log sequence number at the beginning, and the log sequence number
51 at the end.
52 An example command to perform a backup follows:
54 $ xtrabackup --backup --datadir=/var/lib/mysql/ --target-dir=/data/backups/mysql/
56 This takes a backup of /var/lib/mysql and stores it at /data/back‐
58 ups/mysql/. If you specify a relative path, the target directory will
59 be relative to the current directory.
60 During the backup process, you should see a lot of output showing the
62 data files being copied, as well as the log file thread repeatedly
63 scanning the log files and copying from it. Here is an example that
64 shows the log thread scanning the log in the background, and a file
65 copying thread working on the ibdata1 file:
66 >> log scanned up to (3646475465483)
68 >> log scanned up to (3646475517369)
69 >> log scanned up to (3646475581716)
70 >> log scanned up to (3646475636841)
71 >> log scanned up to (3646475718082)
72 >> log scanned up to (3646475988095)
73 >> log scanned up to (3646476048286)
74 >> log scanned up to (3646476102877)
75 >> log scanned up to (3646476140854)
76 [01] Copying /usr/local/mysql/var/ibdata1
77 to /usr/local/mysql/Backups/2011-04-18_21-11-15/ibdata1
78 [01] ...done
79 The last thing you should see is something like the following, where
81 the value of the <LSN> will be a number that depends on your system:
82 xtrabackup: Transaction log of lsn (<SLN>) to (<LSN>) was copied.
84 NOTE:
86 Log copying thread checks the transactional log every second to see
87 if there were any new log records written that need to be copied,
88 but there is a chance that the log copying thread might not be able
89 to keep up with the amount of writes that go to the transactional
90 logs, and will hit an error when the log records are overwritten
91 before they could be read.
92 After the backup is finished, the target directory will contain files
94 such as the following, assuming you have a single InnoDB table
95 test.tbl1 and you are using MySQL's innodb_file_per_table option:
96 /data/backups/mysql/ibdata1
98 /data/backups/mysql/test
99 /data/backups/mysql/test/tbl1.ibd
100 /data/backups/mysql/xtrabackup_checkpoints
101 /data/backups/mysql/xtrabackup_logfile
102 The backup can take a long time, depending on how large the database
104 is. It is safe to cancel at any time, because it does not modify the
105 database.
106 The next step is getting your backup ready to restored: prepar‐
108 ing_the_backup.
109 Preparing the backup
111 After you make a backup with --backup, the next step is to prepare it.
112 The data files are not point-in-time consistent until they've been pre‐
113 pared, because they were copied at different times as the program ran,
114 and they might have been changed while this was happening. If you try
115 to start InnoDB with these data files, it will detect corruption and
116 crash itself to prevent you from running on damaged data. The --prepare
117 step makes the files perfectly consistent at a single instant in time,
118 so you can run InnoDB on them.
119 NOTE:
121 For prepare innobackupex --apply-log should be used which will read
122 InnoDB configuration from backup-my.cnf automatically, or
123 --defaults-file=backup-my.cnf option should be passed to the xtra‐
124 backup binary if it is used for preparing the backup. Otherwise it
125 could lead to incorrect restore because xtrabackup could use wrong
126 configuration options.
127 You can run the prepare operation on any machine; it does not need to
129 be on the originating server or the server to which you intend to
130 restore. You can copy the backup to a utility server and prepare it
131 there, for example.
132 NOTE:
134 You can prepare a backup created with older Percona XtraBackup ver‐
135 sion with a newer one, but not vice versa. Preparing a backup on an
136 unsupported server version should be done with the latest Percona
137 XtraBackup release which supports that server version. For example,
138 if one has a backup of MySQL 5.0 created with Percona XtraBackup
139 1.6, then preparing the backup with Percona XtraBackup 2.2 is not
140 supported, because support for MySQL 5.0 was removed in 2.1.
141 Instead, the latest release in the 2.0 series should be used.
142 During the prepare operation, xtrabackup boots up a kind of modified
144 InnoDB that's embedded inside it (the libraries it was linked against).
145 The modifications are necessary to disable InnoDB's standard safety
146 checks, such as complaining that the log file isn't the right size,
147 which aren't appropriate for working with backups. These modifications
148 are only for the xtrabackup binary; you don't need a modified InnoDB to
149 use xtrabackup for your backups.
150 The prepare step uses this "embedded InnoDB" to perform crash recovery
152 on the copied datafiles, using the copied log file. The prepare step is
153 very simple to use: you simply run xtrabackup with the --prepare option
154 and tell it which directory to prepare, for example, to prepare the
155 backup previously taken,
156 xtrabackup --prepare --target-dir=/data/backups/mysql/
158 When this finishes, you should see an "InnoDB shutdown" with a message
160 such as the following, where again the value of LSN will depend on your
161 system:
162 101107 16:40:15 InnoDB: Shutdown completed; log sequence number <LSN>
164 Your backup is now clean and consistent, and ready to restore. However,
166 you might want to take an extra step to make restores as quick as pos‐
167 sible. This is to prepare the backup a second time. The first time
168 makes the data files perfectly self-consistent, but it doesn't create
169 fresh InnoDB log files. If you restore the backup at this point and
170 start MySQL, it will have to create new log files, which could take a
171 little while, and you might not want to wait for that. If you run
172 --prepare a second time, xtrabackup will create the log files for you,
173 and output status text such as the following, which is abbreviated for
174 clarity. The value of <SIZE> will depend on your MySQL configuration.
175 $ xtrabackup --prepare --target-dir=/data/backups/mysql/
177 xtrabackup: This target seems to be already prepared.
178 xtrabackup: notice: xtrabackup_logfile was already used to '--prepare'.
179 101107 16:54:10 InnoDB: Log file ./ib_logfile0 did not exist: new to be created
180 InnoDB: Setting log file ./ib_logfile0 size to <SIZE> MB
181 InnoDB: Database physically writes the file full: wait...
182 101107 16:54:10 InnoDB: Log file ./ib_logfile1 did not exist: new to be created
183 InnoDB: Setting log file ./ib_logfile1 size to <SIZE> MB
184 InnoDB: Database physically writes the file full: wait...
185 101107 16:54:15 InnoDB: Shutdown completed; log sequence number 1284108
186 All following prepares (third and following) will not change the
188 already prepared data files, you can only see that output says
189 xtrabackup: This target seems to be already prepared.
191 xtrabackup: notice: xtrabackup_logfile was already used to '--prepare'.
192 It is not recommended to interrupt xtrabackup process while preparing
194 backup - it may cause data files corruption and backup will become not
195 usable. Backup validity is not guaranteed if prepare process was inter‐
196 rupted.
197 If you intend the backup to be the basis for further incremental back‐
199 ups, you should use the --apply-log-only option when preparing the
200 backup, or you will not be able to apply incremental backups to it. See
201 the documentation on preparing incremental backups for more details.
202 Restoring a Backup
204 The xtrabackup binary does not have any functionality for restoring a
205 backup. That is up to the user to do. You might use rsync or cp to
206 restore the files. You should check that the restored files have the
207 correct ownership and permissions.
208 NOTE:
210 The datadir must be empty before restoring the backup. Also it's
211 important to note that MySQL server needs to be shut down before
212 restore is performed. You can't restore to a datadir of a running
213 mysqld instance (except when importing a partial backup).
214 Example of the rsync command that can be used to restore the backup can
216 look like this:
217 $ rsync -avrP /data/backup/ /var/lib/mysql/
219 As files' attributes will be preserved, in most cases you will need to
221 change the files' ownership to mysql before starting the database
222 server, as they will be owned by the user who created the backup:
223 $ chown -R mysql:mysql /var/lib/mysql
225 Note that xtrabackup backs up only the InnoDB data. You must separately
227 restore the MySQL system database, MyISAM data, table definition files
228 (.frm files), and everything else necessary to make your database func‐
229 tional -- or innobackupex can do it for you.
230
232 Incremental Backups
233 Both xtrabackup and innobackupex tools supports incremental backups,
234 which means that it can copy only the data that has changed since the
235 last full backup. You can perform many incremental backups between each
236 full backup, so you can set up a backup process such as a full backup
237 once a week and an incremental backup every day, or full backups every
238 day and incremental backups every hour.
239 Incremental backups work because each InnoDB page (usually 16kb in
241 size) contains a log sequence number, or LSN. The LSN is the system
242 version number for the entire database. Each page's LSN shows how
243 recently it was changed. An incremental backup copies each page whose
244 LSN is newer than the previous incremental or full backup's LSN. There
245 are two algorithms in use to find the set of such pages to be copied.
246 The first one, available with all the server types and versions, is to
247 check the page LSN directly by reading all the data pages. The second
248 one, available with Percona Server, is to enable the changed page
249 tracking feature on the server, which will note the pages as they are
250 being changed. This information will be then written out in a compact
251 separate so-called bitmap file. The xtrabackup binary will use that
252 file to read only the data pages it needs for the incremental backup,
253 potentially saving many read requests. The latter algorithm is enabled
254 by default if the xtrabackup binary finds the bitmap file. It is possi‐
255 ble to specify --incremental-force-scan to read all the pages even if
256 the bitmap data is available.
257 Incremental backups do not actually compare the data files to the pre‐
259 vious backup's data files. In fact, you can use --incremental-lsn to
260 perform an incremental backup without even having the previous backup,
261 if you know its LSN. Incremental backups simply read the pages and com‐
262 pare their LSN to the last backup's LSN. You still need a full backup
263 to recover the incremental changes, however; without a full backup to
264 act as a base, the incremental backups are useless.
265 Creating an Incremental Backup
267 To make an incremental backup, begin with a full backup as usual. The
268 xtrabackup binary writes a file called xtrabackup_checkpoints into the
269 backup's target directory. This file contains a line showing the
270 to_lsn, which is the database's LSN at the end of the backup. Create
271 the full backup with a command such as the following:
272 xtrabackup --backup --target-dir=/data/backups/base --datadir=/var/lib/mysql/
274 If you want a usable full backup, use innobackupex since xtrabackup
276 itself won't copy table definitions, triggers, or anything else that's
277 not .ibd.
278 If you look at the xtrabackup_checkpoints file, you should see some
280 contents similar to the following:
281 backup_type = full-backuped
283 from_lsn = 0
284 to_lsn = 1291135
285 Now that you have a full backup, you can make an incremental backup
287 based on it. Use a command such as the following:
288 xtrabackup --backup --target-dir=/data/backups/inc1 \
290 --incremental-basedir=/data/backups/base --datadir=/var/lib/mysql/
291 The /data/backups/inc1/ directory should now contain delta files, such
293 as ibdata1.delta and test/table1.ibd.delta. These represent the changes
294 since the LSN 1291135. If you examine the xtrabackup_checkpoints file
295 in this directory, you should see something similar to the following:
296 backup_type = incremental
298 from_lsn = 1291135
299 to_lsn = 1291340
300 The meaning should be self-evident. It's now possible to use this
302 directory as the base for yet another incremental backup:
303 xtrabackup --backup --target-dir=/data/backups/inc2 \
305 --incremental-basedir=/data/backups/inc1 --datadir=/var/lib/mysql/
306 Preparing the Incremental Backups
308 The --prepare step for incremental backups is not the same as for nor‐
309 mal backups. In normal backups, two types of operations are performed
310 to make the database consistent: committed transactions are replayed
311 from the log file against the data files, and uncommitted transactions
312 are rolled back. You must skip the rollback of uncommitted transactions
313 when preparing a backup, because transactions that were uncommitted at
314 the time of your backup may be in progress, and it's likely that they
315 will be committed in the next incremental backup. You should use the
316 --apply-log-only option to prevent the rollback phase.
317 If you do not use the --apply-log-only option to prevent the rollback
319 phase, then your incremental backups will be useless. After transac‐
320 tions have been rolled back, further incremental backups cannot be
321 applied.
322 Beginning with the full backup you created, you can prepare it, and
324 then apply the incremental differences to it. Recall that you have the
325 following backups:
326 /data/backups/base
328 /data/backups/inc1
329 /data/backups/inc2
330 To prepare the base backup, you need to run --prepare as usual, but
332 prevent the rollback phase:
333 xtrabackup --prepare --apply-log-only --target-dir=/data/backups/base
335 The output should end with some text such as the following:
337 101107 20:49:43 InnoDB: Shutdown completed; log sequence number 1291135
339 The log sequence number should match the to_lsn of the base backup,
341 which you saw previously.
342 This backup is actually safe to restore as-is now, even though the
344 rollback phase has been skipped. If you restore it and start MySQL,
345 InnoDB will detect that the rollback phase was not performed, and it
346 will do that in the background, as it usually does for a crash recovery
347 upon start. It will notify you that the database was not shut down nor‐
348 mally.
349 To apply the first incremental backup to the full backup, you should
351 use the following command:
352 xtrabackup --prepare --apply-log-only --target-dir=/data/backups/base \
354 --incremental-dir=/data/backups/inc1
355 This applies the delta files to the files in /data/backups/base, which
357 rolls them forward in time to the time of the incremental backup. It
358 then applies the redo log as usual to the result. The final data is in
359 /data/backups/base, not in the incremental directory. You should see
360 some output such as the following:
361 incremental backup from 1291135 is enabled.
363 xtrabackup: cd to /data/backups/base/
364 xtrabackup: This target seems to be already prepared.
365 xtrabackup: xtrabackup_logfile detected: size=2097152, start_lsn=(1291340)
366 Applying /data/backups/inc1/ibdata1.delta ...
367 Applying /data/backups/inc1/test/table1.ibd.delta ...
368 .... snip
369 101107 20:56:30 InnoDB: Shutdown completed; log sequence number 1291340
370 Again, the LSN should match what you saw from your earlier inspection
372 of the first incremental backup. If you restore the files from
373 /data/backups/base, you should see the state of the database as of the
374 first incremental backup.
375 Preparing the second incremental backup is a similar process: apply the
377 deltas to the (modified) base backup, and you will roll its data for‐
378 ward in time to the point of the second incremental backup:
379 xtrabackup --prepare --target-dir=/data/backups/base \
381 --incremental-dir=/data/backups/inc2
382 NOTE:
384 --apply-log-only should be used when merging all incrementals except
385 the last one. That's why the previous line doesn't contain the
386 --apply-log-only option. Even if the --apply-log-only was used on
387 the last step, backup would still be consistent but in that case
388 server would perform the rollback phase.
389 If you wish to avoid the notice that InnoDB was not shut down normally,
391 when you have applied the desired deltas to the base backup, you can
392 run --prepare again without disabling the rollback phase.
393 Partial Backups
395 xtrabackup supports taking partial backups when the innodb_file_per_ta‐
396 ble option is enabled. There are three ways to create partial backups:
397 matching the tables' names with a regular expression, providing a list
398 of them in a file or providing a list of databases.
399 WARNING:
401 If any of the matched or listed tables is deleted during the backup,
402 xtrabackup will fail.
403 For the purposes of this manual page, we will assume that there is a
405 database named test which contains tables named t1 and t2.
406 Using the --tables Option
408 The first method is with the --tables option. The option's value is a
409 regular expression that is matched against the fully qualified table‐
410 name, including the database name, in the form databasename.tablename.
411 To back up only tables in the test database, you can use the following
413 command:
414 $ xtrabackup --backup --datadir=/var/lib/mysql --target-dir=/data/backups/ \
416 --tables="^test[.].*"
417 To back up only the table test.t1, you can use the following command:
419 $ xtrabackup --backup --datadir=/var/lib/mysql --target-dir=/data/backups/ \
421 --tables="^test[.]t1"
422 Using the --tables-file Option
424 The --tables-file option specifies a file that can contain multiple ta‐
425 ble names, one table name per line in the file. Only the tables named
426 in the file will be backed up. Names are matched exactly, case-sensi‐
427 tive, with no pattern or regular expression matching. The table names
428 must be fully qualified, in databasename.tablename format.
429 $ echo "mydatabase.mytable" > /tmp/tables.txt
431 $ xtrabackup --backup --tables-file=/tmp/tables.txt
432 Using the --databases and --databases-file options
434 The --databases option accepts a space-separated list of the databases
435 and tables to backup - in the databasename[.tablename] form. The
436 --databases-file option specifies a file that can contain multiple
437 databases and tables in the databasename[.tablename] form, one element
438 name per line in the file. Only named databases and tables will be
439 backed up. Names are matched exactly, case-sensitive, with no pattern
440 or regular expression matching.
441 Preparing the Backup
443 When you use the --prepare option on a partial backup, you will see
444 warnings about tables that don't exist. That is because these tables
445 exist in the data dictionary inside InnoDB, but the corresponding .ibd
446 files don't exist. They were not copied into the backup directory.
447 These tables will be removed from the data dictionary, and when you
448 restore the backup and start InnoDB, they will no longer exist and will
449 not cause any errors or warnings to be printed to the log file.
450 An example of the error message you will see during the prepare phase
452 follows.
453 InnoDB: Reading tablespace information from the .ibd files...
455 101107 22:31:30 InnoDB: Error: table 'test1/t'
456 InnoDB: in InnoDB data dictionary has tablespace id 6,
457 InnoDB: but tablespace with that id or name does not exist. It will be removed from data dictionary.
458 Compact Backups
460 When doing the backup of InnoDB tables it's possible to omit the sec‐
461 ondary index pages. This will make the backups more compact and this
462 way they will take less space on disk. The downside of this is that the
463 backup prepare process takes longer as those secondary indexes need to
464 be recreated. Difference in backup size depends on the size of the sec‐
465 ondary indexes.
466 For example full backup taken without and with the --compact option:
468 #backup size without --compact
470 2.0G xb_backup
471 #backup size taken with --compact option
473 1.4G xb_compact_backup
474 NOTE:
476 Compact backups are not supported for system table space, so in
477 order to work correctly innodb-file-per-table option should be
478 enabled.
479 This feature was introduced in Percona XtraBackup 2.1.
481 Creating Compact Backups
483 To make a compact backup innobackupex needs to be started with the
484 --compact option:
485 $ xtrabackup --backup --compact --target-dir=/data/backups
487 This will create a compact backup in the /data/backups.
489 If you check at the xtrabackup-checkpoints file in the target-dir
491 folder, you should see something like:
492 backup_type = full-backuped
494 from_lsn = 0
495 to_lsn = 2888984349
496 last_lsn = 2888984349
497 compact = 1
498 When --compact wasn't used compact value will be 0. This way it's easy
500 to check if the backup contains the secondary index pages or not.
501 Preparing Compact Backups
503 Preparing the compact require rebuilding the indexes as well. In order
504 to prepare the backup a new option --rebuild-indexes should be used
505 with --apply-logs:
506 $ xtrabackup --prepare --rebuild-indexes /data/backups/
508 Output, beside the standard innobackupex output, should contain the
510 information about indexes being rebuilt, like:
511 [01] Checking if there are indexes to rebuild in table sakila/city (space id: 9)
513 [01] Found index idx_fk_country_id
514 [01] Rebuilding 1 index(es).
515 [01] Checking if there are indexes to rebuild in table sakila/country (space id: 10)
516 [01] Checking if there are indexes to rebuild in table sakila/customer (space id: 11)
517 [01] Found index idx_fk_store_id
518 [01] Found index idx_fk_address_id
519 [01] Found index idx_last_name
520 [01] Rebuilding 3 index(es).
521 Additionally, you can use the --rebuild-threads option to process
523 tables in multiple threads when rebuilding indexes, e.g.:
524 $ xtrabackup --prepare --rebuild-indexes --rebuild-threads=16 /data/backups/
526 In this case Percona XtraBackup will create 16 worker threads with each
528 thread rebuilding indexes for one table at a time. It will also show
529 thread IDs for each message
530 Starting 16 threads to rebuild indexes.
532 [09] Checking if there are indexes to rebuild in table sakila/city (space id: 9)
534 [09] Found index idx_fk_country_id
535 [10] Checking if there are indexes to rebuild in table sakila/country (space id: 10)
536 [11] Checking if there are indexes to rebuild in table sakila/customer (space id: 11)
537 [11] Found index idx_fk_store_id
538 [11] Found index idx_fk_address_id
539 [11] Found index idx_last_name
540 [11] Rebuilding 3 index(es).
541 Since Percona XtraBackup has no information when applying an incremen‐
543 tal backup to a compact full one, on whether there will be more incre‐
544 mental backups applied to it later or not, rebuilding indexes needs to
545 be explicitly requested by a user whenever a full backup with some
546 incremental backups merged is ready to be restored. Rebuilding indexes
547 unconditionally on every incremental backup merge is not an option,
548 since it is an expensive operation.
549 Restoring Compact Backups
551 The xtrabackup binary does not have any functionality for restoring a
552 backup. That is up to the user to do. You might use rsync or cp to
553 restore the files. You should check that the restored files have the
554 correct ownership and permissions.
555 Other Reading
557 · Feature preview: Compact backups in Percona XtraBackup
558
560 Throttling Backups
561 Although xtrabackup does not block your database's operation, any
562 backup can add load to the system being backed up. On systems that do
563 not have much spare I/O capacity, it might be helpful to throttle the
564 rate at which xtrabackup reads and writes data. You can do this with
565 the --throttle option, this option limits the number of I/O operations
566 per second in 1 MB units.
567 Image below shows how throttling works when --throttle =1. [image]
569 In --backup mode, this option limits the number of pairs of
571 read-and-write operations per second that xtrabackup will perform. If
572 you are creating an incremental backup, then the limit is the number of
573 read IO operations per second.
574 By default, there is no throttling, and xtrabackup reads and writes
576 data as quickly as it can. If you set too strict of a limit on the I/O
577 operations, the backup might be so slow that it will never catch up
578 with the transaction logs that InnoDB is writing, so the backup might
579 never complete.
580 Scripting Backups With xtrabackup
582 The xtrabackup tool has several features to enable scripts to control
583 it while they perform related tasks. The innobackupex script is one
584 example, but xtrabackup is easy to control with your own command-line
585 scripts too.
586 Suspending After Copying
588 In backup mode, xtrabackup normally copies the log files in a back‐
589 ground thread, copies the data files in a foreground thread, and then
590 stops the log copying thread and finishes. If you use the --sus‐
591 pend-at-end option, instead of stopping the log thread and finishing,
592 xtrabackup continues to copy the log files, and creates a file in the
593 target directory called xtrabackup_suspended. As long as that file
594 exists, xtrabackup will continue to watch the log files and copy them
595 into the xtrabackup_logfile in the target directory. When the file is
596 removed, xtrabackup will finish copying the log file and exit.
597 This functionality is useful for coordinating the InnoDB data backups
599 with other actions. Perhaps the most obvious is copying the table defi‐
600 nitions (the .frm files) so that the backup can be restored. You can
601 start xtrabackup in the background, wait for the xtrabackup_suspended
602 file to be created, and then copy any other files you need to complete
603 the backup. This is exactly what the innobackupex tool does (it also
604 copies MyISAM data and other files).
605 Generating Meta-Data
607 It is a good idea for the backup to include all the information you
608 need to restore the backup. The xtrabackup tool can print out the con‐
609 tents of a my.cnf file that are needed to restore the data and log
610 files. If you add the --print-param option, it will print out something
611 like the following:
612 # This MySQL options file was generated by XtraBackup.
614 [mysqld]
615 datadir = /data/mysql/
616 innodb_data_home_dir = /data/innodb/
617 innodb_data_file_path = ibdata1:10M:autoextend
618 innodb_log_group_home_dir = /data/innodb-logs/
619 You can redirect this output into a file in the target directory of the
621 backup.
622 Agreeing on the Source Directory
624 It's possible that the presence of a defaults file or other factors
625 could cause xtrabackup to back up data from a different location than
626 you expected. To prevent this, you can use --print-param to ask it
627 where it will be copying data from. You can use the output to ensure
628 that xtrabackup and your script are working on the same dataset.
629 Log Streaming
631 You can instruct xtrabackup to omit copying data files, and simply
632 stream the log file to its standard output instead with --log-stream.
633 This automatically adds the --suspend-at-end option. Your script can
634 then perform tasks such as streaming remote backups by piping the log
635 files into an SSH connection and copying the data files to another
636 server with a tool such as rsync or the xbstream binary.
637 Analyzing Table Statistics
639 The xtrabackup binary can analyze InnoDB data files in read-only mode
640 to give statistics about them. To do this, you should use the --stats
641 option. You can combine this with the --tables option to limit the
642 files to examine. It also uses the --use-memory option.
643 You can perform the analysis on a running server, with some chance of
645 errors due to the data being changed during analysis. Or, you can ana‐
646 lyze a backup copy of the database. Either way, to use the statistics
647 feature, you need a clean copy of the database including correctly
648 sized log files, so you need to execute with --prepare twice to use
649 this functionality on a backup.
650 The result of running on a backup might look like the following:
652 <INDEX STATISTICS>
654 table: test/table1, index: PRIMARY, space id: 12, root page 3
655 estimated statistics in dictionary:
656 key vals: 25265338, leaf pages 497839, size pages 498304
657 real statistics:
658 level 2 pages: pages=1, data=5395 bytes, data/pages=32%
659 level 1 pages: pages=415, data=6471907 bytes, data/pages=95%
660 leaf pages: recs=25958413, pages=497839, data=7492026403 bytes, data/pages=91%
661 This can be interpreted as follows:
663 · The first line simply shows the table and index name and its internal
665 identifiers. If you see an index named GEN_CLUST_INDEX, that is the
666 table's clustered index, automatically created because you did not
667 explicitly create a PRIMARY KEY.
668 · The estimated statistics in dictionary information is similar to the
670 data that's gathered through ANALYZE TABLE inside of InnoDB to be
671 stored as estimated cardinality statistics and passed to the query
672 optimizer.
673 · The real statistics information is the result of scanning the data
675 pages and computing exact information about the index.
676 · The level <X> pages: output means that the line shows information
678 about pages at that level in the index tree. The larger <X> is, the
679 farther it is from the leaf pages, which are level 0. The first line
680 is the root page.
681 · The leaf pages output shows the leaf pages, of course. This is where
683 the table's data is stored.
684 · The external pages: output (not shown) shows large external pages
686 that hold values too long to fit in the row itself, such as long BLOB
687 and TEXT values.
688 · The recs is the real number of records (rows) in leaf pages.
690 · The pages is the page count.
692 · The data is the total size of the data in the pages, in bytes.
694 · The data/pages is calculated as (data / (pages * PAGE_SIZE)) * 100%.
696 It will never reach 100% because of space reserved for page headers
697 and footers.
698 A more detailed example is posted as a MySQL Performance Blog post.
700 Script to Format Output
702 The following script can be used to summarize and tabulate the output
703 of the statistics information:
704 tabulate-xtrabackup-stats.pl
706 #!/usr/bin/perl
708 use strict;
709 use warnings FATAL => 'all';
710 my $script_version = "0.1";
711 my $PG_SIZE = 16_384; # InnoDB defaults to 16k pages, change if needed.
713 my ($cur_idx, $cur_tbl);
714 my (%idx_stats, %tbl_stats);
715 my ($max_tbl_len, $max_idx_len) = (0, 0);
716 while ( my $line = <> ) {
717 if ( my ($t, $i) = $line =~ m/table: (.*), index: (.*), space id:/ ) {
718 $t =~ s!/!.!;
719 $cur_tbl = $t;
720 $cur_idx = $i;
721 if ( length($i) > $max_idx_len ) {
722 $max_idx_len = length($i);
723 }
724 if ( length($t) > $max_tbl_len ) {
725 $max_tbl_len = length($t);
726 }
727 }
728 elsif ( my ($kv, $lp, $sp) = $line =~ m/key vals: (\d+), \D*(\d+), \D*(\d+)/ ) {
729 @{$idx_stats{$cur_tbl}->{$cur_idx}}{qw(est_kv est_lp est_sp)} = ($kv, $lp, $sp);
730 $tbl_stats{$cur_tbl}->{est_kv} += $kv;
731 $tbl_stats{$cur_tbl}->{est_lp} += $lp;
732 $tbl_stats{$cur_tbl}->{est_sp} += $sp;
733 }
734 elsif ( my ($l, $pages, $bytes) = $line =~ m/(?:level (\d+)|leaf) pages:.*pages=(\d+), data=(\d+) bytes/ ) {
735 $l ||= 0;
736 $idx_stats{$cur_tbl}->{$cur_idx}->{real_pages} += $pages;
737 $idx_stats{$cur_tbl}->{$cur_idx}->{real_bytes} += $bytes;
738 $tbl_stats{$cur_tbl}->{real_pages} += $pages;
739 $tbl_stats{$cur_tbl}->{real_bytes} += $bytes;
740 }
741 }
742 my $hdr_fmt = "%${max_tbl_len}s %${max_idx_len}s %9s %10s %10s\n";
744 my @headers = qw(TABLE INDEX TOT_PAGES FREE_PAGES PCT_FULL);
745 printf $hdr_fmt, @headers;
746 my $row_fmt = "%${max_tbl_len}s %${max_idx_len}s %9d %10d %9.1f%%\n";
748 foreach my $t ( sort keys %tbl_stats ) {
749 my $tbl = $tbl_stats{$t};
750 printf $row_fmt, $t, "", $tbl->{est_sp}, $tbl->{est_sp} - $tbl->{real_pages},
751 $tbl->{real_bytes} / ($tbl->{real_pages} * $PG_SIZE) * 100;
752 foreach my $i ( sort keys %{$idx_stats{$t}} ) {
753 my $idx = $idx_stats{$t}->{$i};
754 printf $row_fmt, $t, $i, $idx->{est_sp}, $idx->{est_sp} - $idx->{real_pages},
755 $idx->{real_bytes} / ($idx->{real_pages} * $PG_SIZE) * 100;
756 }
757 }
758 Sample Script Output
760 The output of the above Perl script, when run against the sample shown
761 in the previously mentioned blog post, will appear as follows:
762 TABLE INDEX TOT_PAGES FREE_PAGES PCT_FULL
764 art.link_out104 832383 38561 86.8%
765 art.link_out104 PRIMARY 498304 49 91.9%
766 art.link_out104 domain_id 49600 6230 76.9%
767 art.link_out104 domain_id_2 26495 3339 89.1%
768 art.link_out104 from_message_id 28160 142 96.3%
769 art.link_out104 from_site_id 38848 4874 79.4%
770 art.link_out104 revert_domain 153984 19276 71.4%
771 art.link_out104 site_message 36992 4651 83.4%
772 The columns are the table and index, followed by the total number of
774 pages in that index, the number of pages not actually occupied by data,
775 and the number of bytes of real data as a percentage of the total size
776 of the pages of real data. The first line in the above output, in which
777 the INDEX column is empty, is a summary of the entire table.
778 Working with Binary Logs
780 The xtrabackup binary integrates with information that InnoDB stores in
781 its transaction log about the corresponding binary log position for
782 committed transactions. This enables it to print out the binary log
783 position to which a backup corresponds, so you can use it to set up new
784 replication slaves or perform point-in-time recovery.
785 Finding the Binary Log Position
787 You can find the binary log position corresponding to a backup once the
788 backup has been prepared. This can be done by either running the xtra‐
789 backup with --prepare or innobackupex with --apply-log option. If your
790 backup is from a server with binary logging enabled, xtrabackup will
791 create a file named xtrabackup_binlog_info in the target directory.
792 This file contains the binary log file name and position of the exact
793 point in the binary log to which the prepared backup corresponds.
794 You will also see output similar to the following during the prepare
796 stage:
797 InnoDB: Last MySQL binlog file position 0 3252710, file name ./mysql-bin.000001
799 ... snip ...
800 [notice (again)]
801 If you use binary log and don't use any hack of group commit,
802 the binary log position seems to be:
803 InnoDB: Last MySQL binlog file position 0 3252710, file name ./mysql-bin.000001
804 This output can also be found in the xtrabackup_binlog_pos_innodb file,
806 but it is only correct when no other than XtraDB or InnoDB are used as
807 storage engines.
808 If other storage engines are used (i.e. MyISAM), you should use the
810 xtrabackup_binlog_info file to retrieve the position.
811 The message about hacking group commit refers to an early implementa‐
813 tion of emulated group commit in Percona Server.
814 Point-In-Time Recovery
816 To perform a point-in-time recovery from an xtrabackup backup, you
817 should prepare and restore the backup, and then replay binary logs from
818 the point shown in the xtrabackup_binlog_info file.
819 A more detailed procedure is found here (with innobackupex).
821 Setting Up a New Replication Slave
823 To set up a new replica, you should prepare the backup, and restore it
824 to the data directory of your new replication slave. Then in your
825 CHANGE MASTER TO command, use the binary log filename and position
826 shown in the xtrabackup_binlog_info file to start replication.
827 A more detailed procedure is found in ../howtos/setting_up_replica‐
829 tion.
830 Restoring Individual Tables
832 In server versions prior to 5.6, it is not possible to copy tables
833 between servers by copying the files, even with innodb_file_per_table.
834 However, with Percona XtraBackup, you can export individual tables from
835 any InnoDB database, and import them into Percona Server with XtraDB or
836 MySQL 5.6. (The source doesn't have to be XtraDB or or MySQL 5.6, but
837 the destination does.) This only works on individual .ibd files, and
838 cannot export a table that is not contained in its own .ibd file.
839 Let's see how to export and import the following table:
841 CREATE TABLE export_test (
843 a int(11) DEFAULT NULL
844 ) ENGINE=InnoDB DEFAULT CHARSET=latin1;
845 NOTE:
847 If you're running Percona Server version older than 5.5.10-20.1,
848 variable innodb_expand_import should be used instead of
849 innodb_import_table_from_xtrabackup.
850 Exporting the Table
852 This table should have been created in innodb_file_per_table mode, so
853 after taking a backup as usual with --backup, the .ibd file should
854 exist in the target directory:
855 $ find /data/backups/mysql/ -name export_test.*
857 /data/backups/mysql/test/export_test.ibd
858 when you prepare the backup, add the extra parameter --export to the
860 command. Here is an example:
861 $ xtrabackup --prepare --export --target-dir=/data/backups/mysql/
863 Now you should see a .exp file in the target directory:
865 $ find /data/backups/mysql/ -name export_test.*
867 /data/backups/mysql/test/export_test.exp
868 /data/backups/mysql/test/export_test.ibd
869 /data/backups/mysql/test/export_test.cfg
870 These three files are all you need to import the table into a server
872 running Percona Server with XtraDB or MySQL 5.6.
873 NOTE:
875 MySQL uses .cfg file which contains InnoDB dictionary dump in spe‐
876 cial format. This format is different from the .exp` one which is
877 used in XtraDB for the same purpose. Strictly speaking, a .cfg` file
878 is not required to import a tablespace to MySQL 5.6 or Percona
879 Server 5.6. A tablespace will be imported successfully even if it is
880 from another server, but InnoDB will do schema validation if the
881 corresponding .cfg file is present in the same directory.
882 Importing the Table
884 On the destination server running Percona Server with XtraDB and
885 innodb_import_table_from_xtrabackup option enabled, or MySQL 5.6, cre‐
886 ate a table with the same structure, and then perform the following
887 steps:
888 · Execute ALTER TABLE test.export_test DISCARD TABLESPACE;
890 · If you see the following message, then you must enable inn‐
892 odb_file_per_table and create the table again: ERROR 1030 (HY000):
893 Got error -1 from storage engine
894 · Copy the exported files to the test/ subdirectory of the destination
896 server's data directory
897 · Execute ALTER TABLE test.export_test IMPORT TABLESPACE;
899 The table should now be imported, and you should be able to SELECT from
901 it and see the imported data.
902 NOTE:
904 Persistent statistics for imported tablespace will be empty until
905 you run the ANALYZE TABLE on the imported table. They will be empty
906 because they are stored in the system tables mysql.innodb_ta‐
907 ble_stats and mysql.innodb_index_stats and they aren't updated by
908 server during the import. This is due to upstream bug #72368.
909 LRU dump backup
911 This feature reduces the warm up time by restoring buffer pool state
912 from ib_lru_dump file after restart. Percona XtraBackup discovers
913 ib_lru_dump and backs it up automatically. [image]
914 If the buffer restore option is enabled in my.cnf buffer pool will be
916 in the warm state after backup is restored. To enable this set the
917 variable innodb_buffer_pool_restore_at_startup =1 in Percona Server 5.5
918 or innodb_auto_lru_dump =1 in Percona Server 5.1.
919
921 Implementation Details
922 This page contains notes on various internal aspects of the xtrabackup
923 tool's operation.
924 File Permissions
926 xtrabackup opens the source data files in read-write mode, although it
927 does not modify the files. This means that you must run xtrabackup as a
928 user who has permission to write the data files. The reason for opening
929 the files in read-write mode is that xtrabackup uses the embedded Inn‐
930 oDB libraries to open and read the files, and InnoDB opens them in
931 read-write mode because it normally assumes it is going to write to
932 them.
933 Tuning the OS Buffers
935 Because xtrabackup reads large amounts of data from the filesystem, it
936 uses posix_fadvise() where possible, to instruct the operating system
937 not to try to cache the blocks it reads from disk. Without this hint,
938 the operating system would prefer to cache the blocks, assuming that
939 xtrabackup is likely to need them again, which is not the case. Caching
940 such large files can place pressure on the operating system's virtual
941 memory and cause other processes, such as the database server, to be
942 swapped out. The xtrabackup tool avoids this with the following hint on
943 both the source and destination files:
944 posix_fadvise(file, 0, 0, POSIX_FADV_DONTNEED)
946 In addition, xtrabackup asks the operating system to perform more
948 aggressive read-ahead optimizations on the source files:
949 posix_fadvise(file, 0, 0, POSIX_FADV_SEQUENTIAL)
951 Copying Data Files
953 When copying the data files to the target directory, xtrabackup reads
954 and writes 1MB of data at a time. This is not configurable. When copy‐
955 ing the log file, xtrabackup reads and writes 512 bytes at a time. This
956 is also not possible to configure, and matches InnoDB's behavior (work‐
957 around exists in Percona Server because it has an option to tune inn‐
958 odb_log_block_size for XtraDB, and in that case Percona XtraBackup will
959 match the tuning).
960 After reading from the files, xtrabackup iterates over the 1MB buffer a
962 page at a time, and checks for page corruption on each page with Inn‐
963 oDB's buf_page_is_corrupted() function. If the page is corrupt, it
964 re-reads and retries up to 10 times for each page. It skips this check
965 on the doublewrite buffer.
966 xtrabackup Exit Codes
968 The xtrabackup binary exits with the traditional success value of 0
969 after a backup when no error occurs. If an error occurs during the
970 backup, the exit value is 1.
971 In certain cases, the exit value can be something other than 0 or 1,
973 due to the command-line option code included from the MySQL libraries.
974 An unknown command-line option, for example, will cause an exit code of
975 255.
976
978 The xtrabackup Option Reference
979 This page documents all of the command-line options for the xtrabackup
980 binary.
981 Options
983 --apply-log-only
984 This option causes only the redo stage to be performed when pre‐
985 paring a backup. It is very important for incremental backups.
986 --backup
988 Make a backup and place it in --target-dir. See Creating a
989 backup.
990 --binlog-info
992 This option controls how Percona XtraBackup should retrieve
993 server's binary log coordinates corresponding to the backup.
994 Possible values are OFF, ON, LOCKLESS and AUTO. See the Percona
995 XtraBackup lockless_bin-log manual page for more information
996 --close-files
998 Do not keep files opened. When xtrabackup opens tablespace it
999 normally doesn't close its file handle in order to handle the
1000 DDL operations correctly. However, if the number of tablespaces
1001 is really huge and can not fit into any limit, there is an
1002 option to close file handles once they are no longer accessed.
1003 Percona XtraBackup can produce inconsistent backups with this
1004 option enabled. Use at your own risk.
1005 --compact
1007 Create a compact backup by skipping secondary index pages.
1008 --compress
1010 This option tells xtrabackup to compress all output data,
1011 including the transaction log file and meta data files, using
1012 the specified compression algorithm. The only currently sup‐
1013 ported algorithm is 'quicklz'. The resulting files have the
1014 qpress archive format, i.e. every *.qp file produced by xtra‐
1015 backup is essentially a one-file qpress archive and can be
1016 extracted and uncompressed by the qpress file archiver.
1017 --compress-chunk-size=#
1019 Size of working buffer(s) for compression threads in bytes. The
1020 default value is 64K.
1021 --compress-threads=#
1023 This option specifies the number of worker threads used by xtra‐
1024 backup for parallel data compression. This option defaults to 1.
1025 Parallel compression ('--compress-threads') can be used together
1026 with parallel file copying ('--parallel'). For example, '--par‐
1027 allel=4 --compress --compress-threads=2' will create 4 IO
1028 threads that will read the data and pipe it to 2 compression
1029 threads.
1030 --create-ib-logfile
1032 This option is not currently implemented. To create the InnoDB
1033 log files, you must prepare the backup twice at present.
1034 --datadir=DIRECTORY
1036 The source directory for the backup. This should be the same as
1037 the datadir for your MySQL server, so it should be read from
1038 my.cnf if that exists; otherwise you must specify it on the com‐
1039 mand line.
1040 --defaults-extra-file=[MY.CNF]
1042 Read this file after the global files are read. Must be given as
1043 the first option on the command-line.
1044 --defaults-file=[MY.CNF]
1046 Only read default options from the given file. Must be given as
1047 the first option on the command-line. Must be a real file; it
1048 cannot be a symbolic link.
1049 --defaults-group=GROUP-NAME
1051 This option is to set the group which should be read from the
1052 configuration file. This is used by innobackupex if you use the
1053 --defaults-group option. It is needed for mysqld_multi deploy‐
1054 ments.
1055 --export
1057 Create files necessary for exporting tables. See Restoring Indi‐
1058 vidual Tables.
1059 --extra-lsndir=DIRECTORY
1061 (for --backup): save an extra copy of the xtrabackup_checkpoints
1062 file in this directory.
1063 --incremental-basedir=DIRECTORY
1065 When creating an incremental backup, this is the directory con‐
1066 taining the full backup that is the base dataset for the incre‐
1067 mental backups.
1068 --incremental-dir=DIRECTORY
1070 When preparing an incremental backup, this is the directory
1071 where the incremental backup is combined with the full backup to
1072 make a new full backup.
1073 --incremental-force-scan
1075 When creating an incremental backup, force a full scan of the
1076 data pages in the instance being backuped even if the complete
1077 changed page bitmap data is available.
1078 --incremental-lsn=LSN
1080 When creating an incremental backup, you can specify the log
1081 sequence number (LSN) instead of specifying
1082 --incremental-basedir. For databases created by MySQL and Per‐
1083 cona Server 5.0-series versions, specify the LSN as two 32-bit
1084 integers in high:low format. For databases created in 5.1 and
1085 later, specify the LSN as a single 64-bit integer. ##ATTEN‐
1086 TION##: If a wrong LSN value is specified (a user error which
1087 XtraBackup is unable to detect), the backup will be unusable. Be
1088 careful!
1089 --innodb-log-arch-dir=DIRECTORY
1091 This option is used to specify the directory containing the
1092 archived logs. It can only be used with the xtrabackup --prepare
1093 option.
1094 --innodb-miscellaneous
1096 There is a large group of InnoDB options that are normally read
1097 from the my.cnf configuration file, so that xtrabackup boots up
1098 its embedded InnoDB in the same configuration as your current
1099 server. You normally do not need to specify these explicitly.
1100 These options have the same behavior that they have in InnoDB or
1101 XtraDB. They are as follows:
1102 --innodb-adaptive-hash-index
1104 --innodb-additional-mem-pool-size
1105 --innodb-autoextend-increment
1106 --innodb-buffer-pool-size
1107 --innodb-checksums
1108 --innodb-data-file-path
1109 --innodb-data-home-dir
1110 --innodb-doublewrite-file
1111 --innodb-doublewrite
1112 --innodb-extra-undoslots
1113 --innodb-fast-checksum
1114 --innodb-file-io-threads
1115 --innodb-file-per-table
1116 --innodb-flush-log-at-trx-commit
1117 --innodb-flush-method
1118 --innodb-force-recovery
1119 --innodb-io-capacity
1120 --innodb-lock-wait-timeout
1121 --innodb-log-buffer-size
1122 --innodb-log-files-in-group
1123 --innodb-log-file-size
1124 --innodb-log-group-home-dir
1125 --innodb-max-dirty-pages-pct
1126 --innodb-open-files
1127 --innodb-page-size
1128 --innodb-read-io-threads
1129 --innodb-write-io-threads
1130 --log-copy-interval=#
1132 This option specifies time interval between checks done by log
1133 copying thread in milliseconds (default is 1 second).
1134 --log-stream
1136 Makes xtrabackup not copy data files, and output the contents of
1137 the InnoDB log files to STDOUT until the --suspend-at-end file
1138 is deleted. This option enables --suspend-at-end automatically.
1139 --no-defaults
1141 Don't read default options from any option file. Must be given
1142 as the first option on the command-line.
1143 --databases=#
1145 This option specifies the list of databases and tables that
1146 should be backed up. The option accepts the list of the form
1147 "databasename1[.table_name1] databasename2[.table_name2] . . .".
1148 --databases-file=#
1150 This option specifies the path to the file containing the list
1151 of databases and tables that should be backed up. The file can
1152 contain the list elements of the form databasename1[.ta‐
1153 ble_name1], one element per line.
1154 --parallel=#
1156 This option specifies the number of threads to use to copy mul‐
1157 tiple data files concurrently when creating a backup. The
1158 default value is 1 (i.e., no concurrent transfer).
1159 --prepare
1161 Makes xtrabackup perform recovery on a backup created with
1162 --backup, so that it is ready to use. See preparing a backup.
1163 --print-defaults
1165 Print the program argument list and exit. Must be given as the
1166 first option on the command-line.
1167 --print-param
1169 Makes xtrabackup print out parameters that can be used for copy‐
1170 ing the data files back to their original locations to restore
1171 them. See scripting-xtrabackup.
1172 --rebuild_indexes
1174 Rebuild secondary indexes in InnoDB tables after applying the
1175 log. Only has effect with --prepare.
1176 --rebuild_threads=#
1178 Use this number of threads to rebuild indexes in a compact
1179 backup. Only has effect with --prepare and --rebuild-indexes.
1180 --secure-auth
1182 Refuse client connecting to server if it uses old (pre-4.1.1)
1183 protocol. (Enabled by default; use --skip-secure-auth to dis‐
1184 able.)
1185 --stats
1187 Causes xtrabackup to scan the specified data files and print out
1188 index statistics.
1189 --stream=name
1191 Stream all backup files to the standard output in the specified
1192 format. Currently supported formats are 'xbstream' and 'tar'.
1193 --suspend-at-end
1195 Causes xtrabackup to create a file called xtrabackup_suspended
1196 in the --target-dir. Instead of exiting after copying data
1197 files, xtrabackup continues to copy the log file, and waits
1198 until the xtrabackup_suspended file is deleted. This enables
1199 xtrabackup and other programs to coordinate their work. See
1200 scripting-xtrabackup.
1201 --tables=name
1203 A regular expression against which the full tablename, in
1204 databasename.tablename format, is matched. If the name matches,
1205 the table is backed up. See partial backups.
1206 --tables-file=name
1208 A file containing one table name per line, in database‐
1209 name.tablename format. The backup will be limited to the speci‐
1210 fied tables. See scripting-xtrabackup.
1211 --target-dir=DIRECTORY
1213 This option specifies the destination directory for the backup.
1214 If the directory does not exist, xtrabackup creates it. If the
1215 directory does exist and is empty, xtrabackup will succeed.
1216 xtrabackup will not overwrite existing files, however; it will
1217 fail with operating system error 17, file exists.
1218 If this option is a relative path, it is interpreted as being
1220 relative to the current working directory from which xtrabackup
1221 is executed.
1222 --throttle=#
1224 This option limits --backup to the specified number of
1225 read+write pairs of operations per second. See throttling a
1226 backup.
1227 --tmpdir=name
1229 This option is currently not used for anything except printing
1230 out the correct tmpdir parameter when --print-param is used.
1231 --to-archived-lsn=LSN
1233 This option is used to specify the LSN to which the logs should
1234 be applied when backups are being prepared. It can only be used
1235 with the xtrabackup --prepare option.
1236 --use-memory=#
1238 This option affects how much memory is allocated for preparing a
1239 backup with --prepare, or analyzing statistics with --stats. Its
1240 purpose is similar to innodb_buffer_pool_size. It does not do
1241 the same thing as the similarly named option in Oracle's InnoDB
1242 Hot Backup tool. The default value is 100MB, and if you have
1243 enough available memory, 1GB to 2GB is a good recommended value.
1244 Multiples are supported providing the unit (e.g. 1MB, 1M, 1GB,
1245 1G).
1246 --version
1248 This option prints xtrabackup version and exits.
1249
1251 Percona LLC and/or its affiliates
1252
1254 2009-2016, Percona LLC and/or its affiliates
12552.3.5 Jan 04, 2019 XTRABACKUP(1)