1MK-TABLE-CHECKSUM(1) User Contributed Perl Documentation MK-TABLE-CHECKSUM(1)
2
6 mk-table-checksum - Perform an online replication consistency check, or
7 checksum MySQL tables efficiently on one or many servers.
8
10 Usage: mk-table-checksum [OPTION...] DSN [DSN...]
11 mk-table-checksum checksums MySQL tables efficiently on one or more
13 hosts. Each host is specified as a DSN and missing values are
14 inherited from the first host. If you specify multiple hosts, the
15 first is assumed to be the master.
16 STOP! Are you checksumming a slave(s) against its master? Then be sure
18 to learn what "--replicate" does. It is probably the option you want
19 to use.
20 mk-table-checksum --replicate=mydb.checksum master-host
22 ... time passses, replication catches up ...
23 mk-table-checksum --replicate=mydb.checksum --replicate-check 2 \
24 master-host
25 Or,
27 mk-table-checksum h=host1,u=user,p=password h=host2 ...
29 Or,
31 mk-table-checksum host1 host2 ... hostN | mk-checksum-filter
33 See "SPECIFYING HOSTS" for more on the syntax of the host arguments.
35
37 The following section is included to inform users about the potential
38 risks, whether known or unknown, of using this tool. The two main
39 categories of risks are those created by the nature of the tool (e.g.
40 read-only tools vs. read-write tools) and those created by bugs.
41 mk-table-checksum executes queries that cause the MySQL server to
43 checksum its data. This can cause significant server load. It is
44 read-only unless you use the "--replicate" option, in which case it
45 inserts a small amount of data into the specified table.
46 At the time of this release, we know of no bugs that could cause
48 serious harm to users. There are miscellaneous bugs that might be
49 annoying.
50 The authoritative source for updated information is always the online
52 issue tracking system. Issues that affect this tool will be marked as
53 such. You can see a list of such issues at the following URL:
54 <http://www.maatkit.org/bugs/mk-table-checksum>.
55 See also "BUGS" for more information on filing bugs and getting help.
57
59 mk-table-checksum generates table checksums for MySQL tables, typically
60 useful for verifying your slaves are in sync with the master. The
61 checksums are generated by a query on the server, and there is very
62 little network traffic as a result.
63 Checksums typically take about twice as long as COUNT(*) on very large
65 InnoDB tables in my tests. For smaller tables, COUNT(*) is a good bit
66 faster than the checksums. See "--algorithm" for more details on
67 performance.
68 If you specify more than one server, mk-table-checksum assumes the
70 first server is the master and others are slaves. Checksums are
71 parallelized for speed, forking off a child process for each table.
72 Duplicate server names are ignored, but if you want to checksum a
73 server against itself you can use two different forms of the hostname
74 (for example, "localhost 127.0.0.1", or "h=localhost,P=3306
75 h=localhost,P=3307").
76 If you want to compare the tables in one database to those in another
78 database on the same server, just checksum both databases:
79 mk-table-checksum --databases db1,db2
81 You can then use mk-checksum-filter to compare the results in both
83 databases easily.
84 mk-table-checksum examines table structure only on the first host
86 specified, so if anything differs on the others, it won't notice. It
87 ignores views.
88 The checksums work on MySQL version 3.23.58 through 6.0-alpha. They
90 will not necessarily produce the same values on all versions.
91 Differences in formatting and/or space-padding between 4.1 and 5.0, for
92 example, will cause the checksums to be different.
93
95 mk-table-checksum connects to a theoretically unlimited number of MySQL
96 servers. You specify a list of one or more host definitions on the
97 command line, such as "host1 host2". Each host definition can be just
98 a hostname, or it can be a complex string that specifies connection
99 options as well. You can specify connection options two ways:
100 · Format a host definition in a key=value,key=value form. If an
102 argument on the command line contains the letter '=', mk-table-
103 checksum will parse it into its component parts. Examine the
104 "--help" output for details on the allowed keys.
105 Specifying a list of simple host definitions "host1 host2" is
107 equivalent to the more complicated "h=host1 h=host2" format.
108 · With the command-line options such as "--user" and "--password".
110 These options, if given, apply globally to all host definitions.
111 In addition to specifying connection options this way, mk-table-
113 checksum allows shortcuts. Any options specified for the first host
114 definition on the command line fill in missing values in subsequent
115 ones. Any options that are still missing after this are filled in from
116 the command-line options if possible.
117 In other words, the places you specify connection options have
119 precedence: highest precedence is the option specified directly in the
120 host definition, next is the option specified in the first host
121 definition, and lowest is the command-line option.
122 You can mix simple and complex host definitions and/or command-line
124 arguments. For example, if all your servers except one of your slaves
125 uses a non-standard port number:
126 mk-table-checksum --port 4500 master h=slave1,P=3306 slave2 slave3
128 If you are confused about how mk-table-checksum will connect to your
130 servers, give the "--explain-hosts" option and it will tell you.
131
133 Speed and efficiency are important, because the typical use case is
134 checksumming large amounts of data.
135 "mk-table-checksum" is designed to do very little work itself, and
137 generates very little network traffic aside from inspecting table
138 structures with "SHOW CREATE TABLE". The results of checksum queries
139 are typically 40-character or shorter strings.
140 The MySQL server does the bulk of the work, in the form of the checksum
142 queries. The following benchmarks show the checksum query times for
143 various checksum algorithms. The first two results are simply running
144 "COUNT(col8)" and "CHECKSUM TABLE" on the table. "CHECKSUM TABLE" is
145 just "CRC32" under the hood, but it's implemented inside the storage
146 engine layer instead of at the MySQL layer.
147 ALGORITHM HASH FUNCTION EXTRA TIME
149 ============== ============= ============== =====
150 COUNT(col8) 2.3
151 CHECKSUM TABLE 5.3
152 BIT_XOR FNV_64 12.7
153 ACCUM FNV_64 42.4
154 BIT_XOR MD5 --optimize-xor 80.0
155 ACCUM MD5 87.4
156 BIT_XOR SHA1 --optimize-xor 90.1
157 ACCUM SHA1 101.3
158 BIT_XOR MD5 172.0
159 BIT_XOR SHA1 197.3
160 The tests are entirely CPU-bound. The sample data is an InnoDB table
162 with the following structure:
163 CREATE TABLE test (
165 col1 int NOT NULL,
166 col2 date NOT NULL,
167 col3 int NOT NULL,
168 col4 int NOT NULL,
169 col5 int,
170 col6 decimal(3,1),
171 col7 smallint unsigned NOT NULL,
172 col8 timestamp NOT NULL,
173 PRIMARY KEY (col2, col1),
174 KEY (col7),
175 KEY (col1)
176 ) ENGINE=InnoDB
177 The table has 4303585 rows, 365969408 bytes of data and 173457408 bytes
179 of indexes. The server is a Dell PowerEdge 1800 with dual 32-bit Xeon
180 2.8GHz processors and 2GB of RAM. The tests are fully CPU-bound, and
181 the server is otherwise idle. The results are generally consistent to
182 within a tenth of a second on repeated runs.
183 "CRC32" is the default checksum function to use, and should be enough
185 for most cases. If you need stronger guarantees that your data is
186 identical, you should use one of the other functions.
187
189 The "--algorithm" option allows you to specify which algorithm you
190 would like to use, but it does not guarantee that mk-table-checksum
191 will use this algorithm. mk-table-checksum will ultimately select the
192 best algorithm possible given various factors such as the MySQL version
193 and other command line options.
194 The three basic algorithms in descending order of preference are
196 CHECKSUM, BIT_XOR and ACCUM. CHECKSUM cannot be used if any one of
197 these criteria is true:
198 * L<"--where"> is used.
200 * L<"--since"> is used.
201 * L<"--chunk-size"> is used.
202 * L<"--replicate"> is used.
203 * L<"--count"> is used.
204 * MySQL version less than 4.1.1.
205 The BIT_XOR algorithm also requires MySQL version 4.1.1 or later.
207 After checking these criteria, if the requested "--algorithm" remains
209 then it is used, otherwise the first remaining algorithm with the
210 highest preference is used.
211
213 If you are using this tool to verify your slaves still have the same
214 data as the master, which is why I wrote it, you should read this
215 section.
216 The best way to do this with replication is to use the "--replicate"
218 option. When the queries are finished running on the master and its
219 slaves, you can go to the slaves and issue SQL queries to see if any
220 tables are different from the master. Try the following:
221 SELECT db, tbl, chunk, this_cnt-master_cnt AS cnt_diff,
223 this_crc <> master_crc OR ISNULL(master_crc) <> ISNULL(this_crc)
224 AS crc_diff
225 FROM checksum
226 WHERE master_cnt <> this_cnt OR master_crc <> this_crc
227 OR ISNULL(master_crc) <> ISNULL(this_crc);
228 The "--replicate-check" option can do this query for you. If you can't
230 use this method, try the following:
231 · If your servers are not being written to, you can just run the tool
233 with no further ado:
234 mk-table-checksum server1 server2 ... serverN
236 · If the servers are being written to, you need some way to make sure
238 they are consistent at the moment you run the checksums. For
239 situations other than master-slave replication, you will have to
240 figure this out yourself. You may be able to use the "--where"
241 option with a date or time column to only checksum data that's not
242 recent.
243 · If you are checksumming a master and slaves, you can do a fast
245 parallel checksum and assume the slaves are caught up to the
246 master. In practice, this tends to work well except for tables
247 which are constantly updated. You can use the "--slave-lag" option
248 to see how far behind each slave was when it checksummed a given
249 table. This can help you decide whether to investigate further.
250 · The next most disruptive technique is to lock the table on the
252 master, then take checksums. This should prevent changes from
253 propagating to the slaves. You can just lock on the master (with
254 "--lock"), or you can both lock on the master and wait on the
255 slaves till they reach that point in the master's binlog
256 ("--wait"). Which is better depends on your workload; only you
257 know that.
258 · If you decide to make the checksums on the slaves wait until
260 they're guaranteed to be caught up to the master, the algorithm
261 looks like this:
262 For each table,
264 Master: lock table
265 Master: get pos
266 In parallel,
267 Master: checksum
268 Slave(s): wait for pos, then checksum
269 End
270 Master: unlock table
271 End
272 What I typically do when I'm not using the "--replicate" option is
274 simply run the tool on all servers with no further options. This runs
275 fast, parallel, non-blocking checksums simultaneously. If there are
276 tables that look different, I re-run with "--wait"=600 on the tables in
277 question. This makes the tool lock on the master as explained above.
278
280 Output is to STDOUT, one line per server and table, with header lines
281 for each database. I tried to make the output easy to process with
282 awk. For this reason columns are always present. If there's no value,
283 mk-table-checksum prints 'NULL'.
284 The default is column-aligned output for human readability, but you can
286 change it to tab-separated if you want. Use the "--tab" option for
287 this.
288 Output is unsorted, though all lines for one table should be output
290 together. For speed, all checksums are done in parallel (as much as
291 possible) and may complete out of the order in which they were started.
292 You might want to run them through another script or command-line
293 utility to make sure they are in the order you want. If you pipe the
294 output through mk-checksum-filter, you can sort the output and/or avoid
295 seeing output about tables that have no differences.
296 The columns in the output are as follows. The database, table, and
298 chunk come first so you can sort by them easily (they are the "primary
299 key").
300 Output from "--replicate-check" and "--checksum" are different.
302 DATABASE
304 The database the table is in.
305 TABLE
307 The table name.
308 CHUNK
310 The chunk (see "--chunk-size"). Zero if you are not doing chunked
311 checksums.
312 HOST
314 The server's hostname.
315 ENGINE
317 The table's storage engine.
318 COUNT
320 The table's row count, unless you specified to skip it. If
321 "OVERSIZE" is printed, the chunk was skipped because the actual
322 number of rows was greater than "--chunk-size" times
323 "--chunk-size-limit".
324 CHECKSUM
326 The table's checksum, unless you specified to skip it or the table
327 has no rows. some types of checksums will be 0 if there are no
328 rows; others will print NULL.
329 TIME
331 How long it took to checksum the "CHUNK", not including "WAIT"
332 time. Total checksum time is "WAIT + TIME".
333 WAIT
335 How long the slave waited to catch up to its master before
336 beginning to checksum. "WAIT" is always 0 for the master. See
337 "--wait".
338 STAT
340 The return value of MASTER_POS_WAIT(). "STAT" is always "NULL" for
341 the master.
342 LAG How far the slave lags the master, as reported by SHOW SLAVE
344 STATUS. "LAG" is always "NULL" for the master.
345
347 If you use "--replicate" to store and replicate checksums, you may need
348 to perform maintenance on the replicate table from time to time to
349 remove old checksums. This section describes when checksums in the
350 replicate table are deleted automatically by mk-table-checksum and when
351 you must manually delete them.
352 Before starting, mk-table-checksum calculates chunks for each table,
354 even if "--chunk-size" is not specified (in that case there is one
355 chunk: "1=1"). Then, before checksumming each table, the tool deletes
356 checksum chunks in the replicate table greater than the current number
357 of chunks. For example, if a table is chunked into 100 chunks, 0-99,
358 then mk-table-checksum does:
359 DELETE FROM replicate table WHERE db=? AND tbl=? AND chunk > 99
361 That removes any high-end chunks from previous runs which no longer
363 exist. Currently, this operation cannot be disabled.
364 If you use "--resume", "--resume-replicate", or "--modulo", then you
366 need to be careful that the number of rows in a table does not decrease
367 so much that the number of chunks decreases too, else some checksum
368 chunks may be deleted. The one exception is if only rows at the high
369 end of the range are deleted. In that case, the high-end chunks are
370 deleted and lower chunks remain unchanged. An increasing number of
371 rows or chunks should not cause any adverse affects.
372 Changing the "--chunk-size" between runs with "--resume",
374 "--resume-replicate", or "--modulo" can cause odd or invalid checksums.
375 You should not do this. It won't work with the resume options. With
376 "--modulo", the safest thing to do is manually delete all the rows in
377 the replicate table for the table in question and start over.
378 If the replicate table becomes cluttered with old or invalid checksums
380 and the auto-delete operation is not deleting them, then you will need
381 to manually clean up the replicate table. Alternatively, if you
382 specify "--empty-replicate-table", then the tool deletes every row in
383 the replicate table.
384
386 An exit status of 0 (sometimes also called a return value or return
387 code) indicates success. If there is an error checksumming any table,
388 the exit status is 1.
389 When running "--replicate-check", if any slave has chunks that differ
391 from the master, the exit status is 1.
392
394 If you are using innotop (see <http://code.google.com/p/innotop>),
395 mytop, or another tool to watch currently running MySQL queries, you
396 may see the checksum queries. They look similar to this:
397 REPLACE /*test.test_tbl:'2'/'5'*/ INTO test.checksum(db, ...
399 Since mk-table-checksum's queries run for a long time and tend to be
401 textually very long, and thus won't fit on one screen of these
402 monitoring tools, I've been careful to place a comment at the beginning
403 of the query so you can see what it is and what it's doing. The
404 comment contains the name of the table that's being checksummed, the
405 chunk it is currently checksumming, and how many chunks will be
406 checksummed. In the case above, it is checksumming chunk 2 of 5 in
407 table test.test_tbl.
408
410 "--schema" is restricted to option groups Connection, Filter, Output,
411 Help, Config, Safety.
412 "--empty-replicate-table", "--resume" and "--resume-replicate" are
414 mutually exclusive.
415 This tool accepts additional command-line arguments. Refer to the
417 "SYNOPSIS" and usage information for details.
418 --algorithm
420 type: string
421 Checksum algorithm (ACCUM|CHECKSUM|BIT_XOR).
423 Specifies which checksum algorithm to use. Valid arguments are
425 CHECKSUM, BIT_XOR and ACCUM. The latter two do cryptographic hash
426 checksums. See also "ALGORITHM SELECTION".
427 CHECKSUM is built into MySQL, but has some disadvantages. BIT_XOR
429 and ACCUM are implemented by SQL queries. They use a cryptographic
430 hash of all columns concatenated together with a separator,
431 followed by a bitmap of each nullable column that is NULL
432 (necessary because CONCAT_WS() skips NULL columns).
433 CHECKSUM is the default. This method uses MySQL's built-in
435 CHECKSUM TABLE command, which is a CRC32 behind the scenes. It
436 cannot be used before MySQL 4.1.1, and various options disable it
437 as well. It does not simultaneously count rows; that requires an
438 extra COUNT(*) query. This is a good option when you are using
439 MyISAM tables with live checksums enabled; in this case both the
440 COUNT(*) and CHECKSUM queries will run very quickly.
441 The BIT_XOR algorithm is available for MySQL 4.1.1 and newer. It
443 uses BIT_XOR(), which is order-independent, to reduce all the rows
444 to a single checksum.
445 ACCUM uses a user variable as an accumulator. It reduces each row
447 to a single checksum, which is concatenated with the accumulator
448 and re-checksummed. This technique is order-dependent. If the
449 table has a primary key, it will be used to order the results for
450 consistency; otherwise it's up to chance.
451 The pathological worst case is where identical rows will cancel
453 each other out in the BIT_XOR. In this case you will not be able
454 to distinguish a table full of one value from a table full of
455 another value. The ACCUM algorithm will distinguish them.
456 However, the ACCUM algorithm is order-dependent, so if you have two
458 tables with identical data but the rows are out of order, you'll
459 get different checksums with ACCUM.
460 If a given algorithm won't work for some reason, mk-table-checksum
462 falls back to another. The least common denominator is ACCUM,
463 which works on MySQL 3.23.2 and newer.
464 --arg-table
466 type: string
467 The database.table with arguments for each table to checksum.
469 This table may be named anything you wish. It must contain at
471 least the following columns:
472 CREATE TABLE checksum_args (
474 db char(64) NOT NULL,
475 tbl char(64) NOT NULL,
476 -- other columns as desired
477 PRIMARY KEY (db, tbl)
478 );
479 In addition to the columns shown, it may contain any of the other
481 columns listed here (Note: this list is used by the code,
482 MAGIC_overridable_args):
483 algorithm chunk-column chunk-index chunk-size columns count crc function lock
485 modulo use-index offset optimize-xor chunk-size-limit probability separator
486 save-since single-chunk since since-column sleep sleep-coef trim wait where
487 Each of these columns corresponds to the long form of a command-
489 line option. Each column should be NULL-able. Column names with
490 hyphens should be enclosed in backticks (e.g. `chunk-size`) when
491 the table is created. The data type does not matter, but it's
492 suggested you use a sensible data type to prevent garbage data.
493 When "mk-table-checksum" checksums a table, it will look for a
495 matching entry in this table. Any column that has a defined value
496 will override the corresponding command-line argument for the table
497 being currently processed. In this way it is possible to specify
498 custom command-line arguments for any table.
499 If you add columns to the table that aren't in the above list of
501 allowable columns, it's an error. The exceptions are "db", "tbl",
502 and "ts". The "ts" column can be used as a timestamp for easy
503 visibility into the last time the "since" column was updated with
504 "--save-since".
505 This table is assumed to be located on the first server given on
507 the command-line.
508 --ask-pass
510 group: Connection
511 Prompt for a password when connecting to MySQL.
513 --check-interval
515 type: time; group: Throttle; default: 1s
516 How often to check for slave lag if "--check-slave-lag" is given.
518 --[no]check-replication-filters
520 default: yes; group: Safety
521 Do not "--replicate" if any replication filters are set. When
523 --replicate is specified, mk-table-checksum tries to detect slaves
524 and look for options that filter replication, such as
525 binlog_ignore_db and replicate_do_db. If it finds any such
526 filters, it aborts with an error. Replication filtering makes it
527 impossible to be sure that the checksum queries won't break
528 replication or simply fail to replicate. If you are sure that it's
529 OK to run the checksum queries, you can negate this option to
530 disable the checks. See also "--replicate-database".
531 --check-slave-lag
533 type: DSN; group: Throttle
534 Pause checksumming until the specified slave's lag is less than
536 "--max-lag".
537 If this option is specified and "--throttle-method" is set to
539 "slavelag" then "--throttle-method" only checks this slave.
540 --checksum
542 group: Output
543 Print checksums and table names in the style of md5sum (disables
545 "--[no]count").
546 Makes the output behave more like the output of "md5sum". The
548 checksum is first on the line, followed by the host, database,
549 table, and chunk number, concatenated with dots.
550 --chunk-column
552 type: string
553 Prefer this column for dividing tables into chunks. By default,
555 mk-table-checksum chooses the first suitable column for each table,
556 preferring to use the primary key. This option lets you specify a
557 preferred column, which mk-table-checksum uses if it exists in the
558 table and is chunkable. If not, then mk-table-checksum will revert
559 to its default behavior. Be careful when using this option; a poor
560 choice could cause bad performance. This is probably best to use
561 when you are checksumming only a single table, not an entire
562 server. See also "--chunk-index".
563 --chunk-index
565 type: string
566 Prefer this index for chunking tables. By default, mk-table-
568 checksum chooses an appropriate index for the "--chunk-column"
569 (even if it chooses the chunk column automatically). This option
570 lets you specify the index you prefer. If the index doesn't exist,
571 then mk-table-checksum will fall back to its default behavior. mk-
572 table-checksum adds the index to the checksum SQL statements in a
573 "FORCE INDEX" clause. Be careful when using this option; a poor
574 choice of index could cause bad performance. This is probably best
575 to use when you are checksumming only a single table, not an entire
576 server.
577 --chunk-range
579 type: string; default: open
580 Set which ends of the chunk range are open or closed. Possible
582 values are one of MAGIC_chunk_range:
583 VALUE OPENS/CLOSES
585 ========== ======================
586 open Both ends are open
587 openclosed Low end open, high end closed
588 By default mk-table-checksum uses an open range of chunks like:
590 `id` < '10'
592 `id` >= '10' AND < '20'
593 `id` >= '20'
594 That range is open because the last chunk selects any row with id
596 greater than (or equal to) 20. An open range can be a problem in
597 cases where a lot of new rows are inserted with IDs greater than 20
598 while mk-table-checksumming is running because the final open-ended
599 chunk will select all the newly inserted rows. (The less common
600 case of inserting rows with IDs less than 10 would require a
601 "closedopen" range but that is not currently implemented.)
602 Specifying "openclosed" will cause the final chunk to be closed
603 like:
604 `id` >= '20' AND `id` <= N
606 N is the "MAX(`id`)" that mk-table-checksum used when it first
608 chunked the rows. Therefore, it will only chunk the range of rows
609 that existed when the tool started and not any newly inserted rows
610 (unless those rows happen to be inserted with IDs less than N).
611 See also "--chunk-size-limit".
613 --chunk-size
615 type: string
616 Approximate number of rows or size of data to checksum at a time.
618 Allowable suffixes are k, M, G. Disallows "--algorithm CHECKSUM".
619 If you specify a chunk size, mk-table-checksum will try to find an
621 index that will let it split the table into ranges of approximately
622 "--chunk-size" rows, based on the table's index statistics.
623 Currently only numeric and date types can be chunked.
624 If the table is chunkable, mk-table-checksum will checksum each
626 range separately with parameters in the checksum query's WHERE
627 clause. If mk-table-checksum cannot find a suitable index, it will
628 do the entire table in one chunk as though you had not specified
629 "--chunk-size" at all. Each table is handled individually, so some
630 tables may be chunked and others not.
631 The chunks will be approximately sized, and depending on the
633 distribution of values in the indexed column, some chunks may be
634 larger than the value you specify.
635 If you specify a suffix (one of k, M or G), the parameter is
637 treated as a data size rather than a number of rows. The output of
638 SHOW TABLE STATUS is then used to estimate the amount of data the
639 table contains, and convert that to a number of rows.
640 --chunk-size-limit
642 type: float; default: 2.0; group: Safety
643 Do not checksum chunks with this many times more rows than
645 "--chunk-size".
646 When "--chunk-size" is given it specifies an ideal size for each
648 chunk of a chunkable table (in rows; size values are converted to
649 rows). Before checksumming each chunk, mk-table-checksum checks
650 how many rows are in the chunk with EXPLAIN. If the number of rows
651 reported by EXPLAIN is this many times greater than "--chunk-size",
652 then the chunk is skipped and "OVERSIZE" is printed for the "COUNT"
653 column of the "OUTPUT".
654 For example, if you specify "--chunk-size" 100 and a chunk has 150
656 rows, then it is checksummed with the default "--chunk-size-limit"
657 value 2.0 because 150 is less than 100 * 2.0. But if the chunk has
658 205 rows, then it is not checksummed because 205 is greater than
659 100 * 2.0.
660 The minimum value for this option is 1 which means that no chunk
662 can be any larger than "--chunk-size". You probably don't want to
663 specify 1 because rows reported by EXPLAIN are estimates which can
664 be greater than or less than the real number of rows in the chunk.
665 If too many chunks are skipped because they are oversize, you might
666 want to specify a value larger than 2.
667 You can disable oversize chunk checking by specifying
669 "--chunk-size-limit" 0.
670 See also "--unchunkable-tables".
672 --columns
674 short form: -c; type: array; group: Filter
675 Checksum only this comma-separated list of columns.
677 --config
679 type: Array; group: Config
680 Read this comma-separated list of config files; if specified, this
682 must be the first option on the command line.
683 --[no]count
685 Count rows in tables. This is built into ACCUM and BIT_XOR, but
686 requires an extra query for CHECKSUM.
687 This is disabled by default to avoid an extra COUNT(*) query when
689 "--algorithm" is CHECKSUM. If you have only MyISAM tables and live
690 checksums are enabled, both CHECKSUM and COUNT will be very fast,
691 but otherwise you may want to use one of the other algorithms.
692 --[no]crc
694 default: yes
695 Do a CRC (checksum) of tables.
697 Take the checksum of the rows as well as their count. This is
699 enabled by default. If you disable it, you'll just get COUNT(*)
700 queries.
701 --create-replicate-table
703 Create the replicate table given by "--replicate" if it does not
704 exist.
705 Normally, if the replicate table given by "--replicate" does not
707 exist, "mk-table-checksum" will die. With this option, however,
708 "mk-table-checksum" will create the replicate table for you, using
709 the database.table name given to "--replicate".
710 The structure of the replicate table is the same as the suggested
712 table mentioned in "--replicate". Note that since ENGINE is not
713 specified, the replicate table will use the server's default
714 storage engine. If you want to use a different engine, you need to
715 create the table yourself.
716 --databases
718 short form: -d; type: hash; group: Filter
719 Only checksum this comma-separated list of databases.
721 --databases-regex
723 type: string
724 Only checksum databases whose names match this Perl regex.
726 --defaults-file
728 short form: -F; type: string; group: Connection
729 Only read mysql options from the given file. You must give an
731 absolute pathname.
732 --empty-replicate-table
734 DELETE all rows in the "--replicate" table before starting.
735 Issues a DELETE against the table given by "--replicate" before
737 beginning work. Ignored if "--replicate" is not specified. This
738 can be useful to remove entries related to tables that no longer
739 exist, or just to clean out the results of a previous run.
740 If you want to delete entries for specific databases or tables you
742 must do this manually.
743 --engines
745 short form: -e; type: hash; group: Filter
746 Do only this comma-separated list of storage engines.
748 --explain
750 group: Output
751 Show, but do not execute, checksum queries (disables
753 "--empty-replicate-table").
754 --explain-hosts
756 group: Help
757 Print connection information and exit.
759 Print out a list of hosts to which mk-table-checksum will connect,
761 with all the various connection options, and exit. See "SPECIFYING
762 HOSTS".
763 --float-precision
765 type: int
766 Precision for "FLOAT" and "DOUBLE" number-to-string conversion.
768 Causes FLOAT and DOUBLE values to be rounded to the specified
769 number of digits after the decimal point, with the ROUND() function
770 in MySQL. This can help avoid checksum mismatches due to different
771 floating-point representations of the same values on different
772 MySQL versions and hardware. The default is no rounding; the
773 values are converted to strings by the CONCAT() function, and MySQL
774 chooses the string representation. If you specify a value of 2,
775 for example, then the values 1.008 and 1.009 will be rounded to
776 1.01, and will checksum as equal.
777 --function
779 type: string
780 Hash function for checksums (FNV1A_64, MURMUR_HASH, SHA1, MD5,
782 CRC32, etc).
783 You can use this option to choose the cryptographic hash function
785 used for "--algorithm"=ACCUM or "--algorithm"=BIT_XOR. The default
786 is to use "CRC32", but "MD5" and "SHA1" also work, and you can use
787 your own function, such as a compiled UDF, if you wish. Whatever
788 function you specify is run in SQL, not in Perl, so it must be
789 available to MySQL.
790 The "FNV1A_64" UDF mentioned in the benchmarks is much faster than
792 "MD5". The C++ source code is distributed with Maatkit. It is
793 very simple to compile and install; look at the header in the
794 source code for instructions. If it is installed, it is preferred
795 over "MD5". You can also use the MURMUR_HASH function if you
796 compile and install that as a UDF; the source is also distributed
797 with Maatkit, and it is faster and has better distribution than
798 FNV1A_64.
799 --help
801 group: Help
802 Show help and exit.
804 --ignore-columns
806 type: Hash; group: Filter
807 Ignore this comma-separated list of columns when calculating the
809 checksum.
810 This option only affects the checksum when using the ACCUM or
812 BIT_XOR "--algorithm".
813 --ignore-databases
815 type: Hash; group: Filter
816 Ignore this comma-separated list of databases.
818 --ignore-databases-regex
820 type: string
821 Ignore databases whose names match this Perl regex.
823 --ignore-engines
825 type: Hash; default: FEDERATED,MRG_MyISAM; group: Filter
826 Ignore this comma-separated list of storage engines.
828 --ignore-tables
830 type: Hash; group: Filter
831 Ignore this comma-separated list of tables.
833 Table names may be qualified with the database name.
835 --ignore-tables-regex
837 type: string
838 Ignore tables whose names match the Perl regex.
840 --lock
842 Lock on master until done on slaves (implies "--slave-lag").
843 This option can help you to get a consistent read on a master and
845 many slaves. If you specify this option, mk-table-checksum will
846 lock the table on the first server on the command line, which it
847 assumes to be the master. It will keep this lock until the
848 checksums complete on the other servers.
849 This option isn't very useful by itself, so you probably want to
851 use "--wait" instead.
852 Note: if you're checksumming a slave against its master, you should
854 use "--replicate". In that case, there's no need for locking,
855 waiting, or any of that.
856 --max-lag
858 type: time; group: Throttle; default: 1s
859 Suspend checksumming if the slave given by "--check-slave-lag"
861 lags.
862 This option causes mk-table-checksum to look at the slave every
864 time it's about to checksum a chunk. If the slave's lag is greater
865 than the option's value, or if the slave isn't running (so its lag
866 is NULL), mk-table-checksum sleeps for "--check-interval" seconds
867 and then looks at the lag again. It repeats until the slave is
868 caught up, then proceeds to checksum the chunk.
869 This option is useful to let you checksum data as fast as the
871 slaves can handle it, assuming the slave you directed mk-table-
872 checksum to monitor is representative of all the slaves that may be
873 replicating from this server. It should eliminate the need for
874 "--sleep" or "--sleep-coef".
875 --modulo
877 type: int
878 Do only every Nth chunk on chunked tables.
880 This option lets you checksum only some chunks of the table. This
882 is a useful alternative to "--probability" when you want to be sure
883 you get full coverage in some specified number of runs; for
884 example, you can do only every 7th chunk, and then use "--offset"
885 to rotate the modulo every day of the week.
886 Just like with "--probability", a table that cannot be chunked is
888 done every time.
889 --offset
891 type: string; default: 0
892 Modulo offset expression for use with "--modulo".
894 The argument may be an SQL expression, such as "WEEKDAY(NOW())"
896 (which returns a number from 0 through 6). The argument is
897 evaluated by MySQL. The result is used as follows: if chunk_num %
898 "--modulo" == "--offset", the chunk will be checksummed.
899 --[no]optimize-xor
901 default: yes
902 Optimize BIT_XOR with user variables.
904 This option specifies to use user variables to reduce the number of
906 times each row must be passed through the cryptographic hash
907 function when you are using the BIT_XOR algorithm.
908 With the optimization, the queries look like this in pseudo-code:
910 SELECT CONCAT(
912 BIT_XOR(SLICE_OF(@user_variable)),
913 BIT_XOR(SLICE_OF(@user_variable)),
914 ...
915 BIT_XOR(SLICE_OF(@user_variable := HASH(col1, col2... colN))));
916 The exact positioning of user variables and calls to the hash
918 function is determined dynamically, and will vary between MySQL
919 versions. Without the optimization, it looks like this:
920 SELECT CONCAT(
922 BIT_XOR(SLICE_OF(MD5(col1, col2... colN))),
923 BIT_XOR(SLICE_OF(MD5(col1, col2... colN))),
924 ...
925 BIT_XOR(SLICE_OF(MD5(col1, col2... colN))));
926 The difference is the number of times all the columns must be
928 mashed together and fed through the hash function. If you are
929 checksumming really large columns, such as BLOB or TEXT columns,
930 this might make a big difference.
931 --password
933 short form: -p; type: string; group: Connection
934 Password to use when connecting.
936 --pid
938 type: string
939 Create the given PID file. The file contains the process ID of the
941 script. The PID file is removed when the script exits. Before
942 starting, the script checks if the PID file already exists. If it
943 does not, then the script creates and writes its own PID to it. If
944 it does, then the script checks the following: if the file contains
945 a PID and a process is running with that PID, then the script dies;
946 or, if there is no process running with that PID, then the script
947 overwrites the file with its own PID and starts; else, if the file
948 contains no PID, then the script dies.
949 --port
951 short form: -P; type: int; group: Connection
952 Port number to use for connection.
954 --probability
956 type: int; default: 100
957 Checksums will be run with this percent probability.
959 This is an integer between 1 and 100. If 100, every chunk of every
961 table will certainly be checksummed. If less than that, there is a
962 chance that some chunks of some tables will be skipped. This is
963 useful for routine jobs designed to randomly sample bits of tables
964 without checksumming the whole server. By default, if a table is
965 not chunkable, it will be checksummed every time even when the
966 probability is less than 100. You can override this with
967 "--single-chunk".
968 See also "--modulo".
970 --progress
972 type: array; default: time,30
973 Print progress reports to STDERR. Currently, this feature is only
975 for when "--throttle-method" waits for slaves to catch up.
976 The value is a comma-separated list with two parts. The first part
978 can be percentage, time, or iterations; the second part specifies
979 how often an update should be printed, in percentage, seconds, or
980 number of iterations.
981 --quiet
983 short form: -q; group: Output
984 Do not print checksum results.
986 --recheck
988 Re-checksum chunks that "--replicate-check" found to be different.
989 --recurse
991 type: int; group: Throttle
992 Number of levels to recurse in the hierarchy when discovering
994 slaves. Default is infinite.
995 See "--recursion-method".
997 --recursion-method
999 type: string
1000 Preferred recursion method for discovering slaves.
1002 Possible methods are:
1004 METHOD USES
1006 =========== ================
1007 processlist SHOW PROCESSLIST
1008 hosts SHOW SLAVE HOSTS
1009 The processlist method is preferred because SHOW SLAVE HOSTS is not
1011 reliable. However, the hosts method is required if the server uses
1012 a non-standard port (not 3306). Usually mk-table-checksum does the
1013 right thing and finds the slaves, but you may give a preferred
1014 method and it will be used first. If it doesn't find any slaves,
1015 the other methods will be tried.
1016 --replicate
1018 type: string
1019 Replicate checksums to slaves (disallows --algorithm CHECKSUM).
1021 This option enables a completely different checksum strategy for a
1023 consistent, lock-free checksum across a master and its slaves.
1024 Instead of running the checksum queries on each server, you run
1025 them only on the master. You specify a table, fully qualified in
1026 db.table format, to insert the results into. The checksum queries
1027 will insert directly into the table, so they will be replicated
1028 through the binlog to the slaves.
1029 When the queries are finished replicating, you can run a simple
1031 query on each slave to see which tables have differences from the
1032 master. With the "--replicate-check" option, mk-table-checksum can
1033 run the query for you to make it even easier. See "CONSISTENT
1034 CHECKSUMS" for details.
1035 If you find tables that have differences, you can use the chunk
1037 boundaries in a WHERE clause with mk-table-sync to help repair them
1038 more efficiently. See mk-table-sync for details.
1039 The table must have at least these columns: db, tbl, chunk,
1041 boundaries, this_crc, master_crc, this_cnt, master_cnt. The table
1042 may be named anything you wish. Here is a suggested table
1043 structure, which is automatically used for
1044 "--create-replicate-table" (MAGIC_create_replicate):
1045 CREATE TABLE checksum (
1047 db char(64) NOT NULL,
1048 tbl char(64) NOT NULL,
1049 chunk int NOT NULL,
1050 boundaries char(100) NOT NULL,
1051 this_crc char(40) NOT NULL,
1052 this_cnt int NOT NULL,
1053 master_crc char(40) NULL,
1054 master_cnt int NULL,
1055 ts timestamp NOT NULL,
1056 PRIMARY KEY (db, tbl, chunk)
1057 );
1058 Be sure to choose an appropriate storage engine for the checksum
1060 table. If you are checksumming InnoDB tables, for instance, a
1061 deadlock will break replication if the checksum table is non-
1062 transactional, because the transaction will still be written to the
1063 binlog. It will then replay without a deadlock on the slave and
1064 break replication with "different error on master and slave." This
1065 is not a problem with mk-table-checksum, it's a problem with MySQL
1066 replication, and you can read more about it in the MySQL manual.
1067 This works only with statement-based replication (mk-table-checksum
1069 will switch the binlog format to STATEMENT for the duration of the
1070 session if your server uses row-based replication).
1071 In contrast to running the tool against multiple servers at once,
1073 using this option eliminates the complexities of synchronizing
1074 checksum queries across multiple servers, which normally requires
1075 locking and unlocking, waiting for master binlog positions, and so
1076 on. Thus, it disables "--lock", "--wait", and "--slave-lag" (but
1077 not "--check-slave-lag", which is a way to throttle the execution
1078 speed).
1079 The checksum queries actually do a REPLACE into this table, so
1081 existing rows need not be removed before running. However, you may
1082 wish to do this anyway to remove rows related to tables that don't
1083 exist anymore. The "--empty-replicate-table" option does this for
1084 you.
1085 Since the table must be qualified with a database (e.g.
1087 "db.checksums"), mk-table-checksum will only USE this database.
1088 This may be important if any replication options are set because it
1089 could affect whether or not changes to the table are replicated.
1090 If the slaves have any --replicate-do-X or --replicate-ignore-X
1092 options, you should be careful not to checksum any databases or
1093 tables that exist on the master and not the slaves. Changes to
1094 such tables may not normally be executed on the slaves because of
1095 the --replicate options, but the checksum queries modify the
1096 contents of the table that stores the checksums, not the tables
1097 whose data you are checksumming. Therefore, these queries will be
1098 executed on the slave, and if the table or database you're
1099 checksumming does not exist, the queries will cause replication to
1100 fail. For more information on replication rules, see
1101 <http://dev.mysql.com/doc/en/replication-rules.html>.
1102 The table specified by "--replicate" will never be checksummed
1104 itself.
1105 --replicate-check
1107 type: int
1108 Check results in "--replicate" table, to the specified depth. You
1110 must use this after you run the tool normally; it skips the
1111 checksum step and only checks results.
1112 It recursively finds differences recorded in the table given by
1114 "--replicate". It recurses to the depth you specify: 0 is no
1115 recursion (check only the server you specify), 1 is check the
1116 server and its slaves, 2 is check the slaves of its slaves, and so
1117 on.
1118 It finds differences by running the query shown in "CONSISTENT
1120 CHECKSUMS", and prints results, then exits after printing. This is
1121 just a convenient way of running the query so you don't have to do
1122 it manually.
1123 The output is one informational line per slave host, followed by
1125 the results of the query, if any. If "--quiet" is specified, there
1126 is no output. If there are no differences between the master and
1127 any slave, there is no output. If any slave has chunks that differ
1128 from the master, mk-table-checksum's exit status is 1; otherwise it
1129 is 0.
1130 This option makes "mk-table-checksum" look for slaves by running
1132 "SHOW PROCESSLIST". If it finds connections that appear to be from
1133 slaves, it derives connection information for each slave with the
1134 same default-and-override method described in "SPECIFYING HOSTS".
1135 If "SHOW PROCESSLIST" doesn't return any rows, "mk-table-checksum"
1137 looks at "SHOW SLAVE HOSTS" instead. The host and port, and user
1138 and password if available, from "SHOW SLAVE HOSTS" are combined
1139 into a DSN and used as the argument. This requires slaves to be
1140 configured with "report-host", "report-port" and so on.
1141 This requires the @@SERVER_ID system variable, so it works only on
1143 MySQL 3.23.26 or newer.
1144 --replicate-database
1146 type: string
1147 "USE" only this database with "--replicate". By default, mk-table-
1149 checksum executes USE to set its default database to the database
1150 that contains the table it's currently working on. It changes its
1151 default database as it works on different tables. This is is a
1152 best effort to avoid problems with replication filters such as
1153 binlog_ignore_db and replicate_ignore_db. However, replication
1154 filters can create a situation where there simply is no one right
1155 way to do things. Some statements might not be replicated, and
1156 others might cause replication to fail on the slaves. In such
1157 cases, it is up to the user to specify a safe default database.
1158 This option specifies a default database that mk-table-checksum
1159 selects with USE, and never changes afterwards. See also
1160 <L"--[no]check-replication-filters">.
1161 --resume
1163 type: string
1164 Resume checksum using given output file from a previously
1166 interrupted run.
1167 The given output file should be the literal output from a previous
1169 run of "mk-table-checksum". For example:
1170 mk-table-checksum host1 host2 -C 100 > checksum_results.txt
1172 mk-table-checksum host1 host2 -C 100 --resume checksum_results.txt
1173 The command line options given to the first run and the resumed run
1175 must be identical (except, of course, for --resume). If they are
1176 not, the result will be unpredictable and probably wrong.
1177 "--resume" does not work with "--replicate"; for that, use
1179 "--resume-replicate".
1180 --resume-replicate
1182 Resume "--replicate".
1183 This option resumes a previous checksum operation using
1185 "--replicate". It is like "--resume" but does not require an
1186 output file. Instead, it uses the checksum table given to
1187 "--replicate" to determine where to resume the checksum operation.
1188 --save-since
1190 When "--arg-table" and "--since" are given, save the current
1191 "--since" value into that table's "since" column after
1192 checksumming. In this way you can incrementally checksum tables by
1193 starting where the last one finished.
1194 The value to be saved could be the current timestamp, or it could
1196 be the maximum existing value of the column given by
1197 "--since-column". It depends on what options are in effect. See
1198 the description of "--since" to see how timestamps are different
1199 from ordinary values.
1200 --schema
1202 Checksum "SHOW CREATE TABLE" instead of table data.
1203 --separator
1205 type: string; default: #
1206 The separator character used for CONCAT_WS().
1208 This character is used to join the values of columns when
1210 checksumming with "--algorithm" of BIT_XOR or ACCUM.
1211 --set-vars
1213 type: string; default: wait_timeout=10000; group: Connection
1214 Set these MySQL variables. Immediately after connecting to MySQL,
1216 this string will be appended to SET and executed.
1217 --since
1219 type: string
1220 Checksum only data newer than this value.
1222 If the table is chunk-able or nibble-able, this value will apply to
1224 the first column of the chunked or nibbled index.
1225 This is not too different to "--where", but instead of universally
1227 applying a WHERE clause to every table, it selectively finds the
1228 right column to use and applies it only if such a column is found.
1229 See also "--since-column".
1230 The argument may be an expression, which is evaluated by MySQL.
1232 For example, you can specify "CURRENT_DATE - INTERVAL 7 DAY" to get
1233 the date of one week ago.
1234 A special bit of extra magic: if the value is temporal (looks like
1236 a date or datetime), then the table is checksummed only if the
1237 create time (or last modified time, for tables that report the last
1238 modified time, such as MyISAM tables) is newer than the value. In
1239 this sense it's not applied as a WHERE clause at all.
1240 --since-column
1242 type: string
1243 The column name to be used for "--since".
1245 The default is for the tool to choose the best one automatically.
1247 If you specify a value, that will be used if possible; otherwise
1248 the best auto-determined one; otherwise none. If the column
1249 doesn't exist in the table, it is just ignored.
1250 --single-chunk
1252 Permit skipping with "--probability" if there is only one chunk.
1253 Normally, if a table isn't split into many chunks, it will always
1255 be checksummed regardless of "--probability". This setting lets
1256 the probabilistic behavior apply to tables that aren't divided into
1257 chunks.
1258 --slave-lag
1260 group: Output
1261 Report replication delay on the slaves.
1263 If this option is enabled, the output will show how many seconds
1265 behind the master each slave is. This can be useful when you want
1266 a fast, parallel, non-blocking checksum, and you know your slaves
1267 might be delayed relative to the master. You can inspect the
1268 results and make an educated guess whether any discrepancies on the
1269 slave are due to replication delay instead of corrupt data.
1270 If you're using "--replicate", a slave that is delayed relative to
1272 the master does not invalidate the correctness of the results, so
1273 this option is disabled.
1274 --sleep
1276 type: int; group: Throttle
1277 Sleep time between checksums.
1279 If this option is specified, mk-table-checksum will sleep the
1281 specified number of seconds between checksums. That is, it will
1282 sleep between every table, and if you specify "--chunk-size", it
1283 will also sleep between chunks.
1284 This is a very crude way to throttle checksumming; see
1286 "--sleep-coef" and "--check-slave-lag" for techniques that permit
1287 greater control.
1288 --sleep-coef
1290 type: float; group: Throttle
1291 Calculate "--sleep" as a multiple of the last checksum time.
1293 If this option is specified, mk-table-checksum will sleep the
1295 amount of time elapsed during the previous checksum, multiplied by
1296 the specified coefficient. This option is ignored if "--sleep" is
1297 specified.
1298 This is a slightly more sophisticated way to throttle checksum
1300 speed: sleep a varying amount of time between chunks, depending on
1301 how long the chunks are taking. Even better is to use
1302 "--check-slave-lag" if you're checksumming master/slave
1303 replication.
1304 --socket
1306 short form: -S; type: string; group: Connection
1307 Socket file to use for connection.
1309 --tab
1311 group: Output
1312 Print tab-separated output, not column-aligned output.
1314 --tables
1316 short form: -t; type: hash; group: Filter
1317 Do only this comma-separated list of tables.
1319 Table names may be qualified with the database name.
1321 --tables-regex
1323 type: string
1324 Only checksum tables whose names match this Perl regex.
1326 --throttle-method
1328 type: string; default: none; group: Throttle
1329 Throttle checksumming when doing "--replicate".
1331 At present there is only one method: "slavelag". When
1333 "--replicate" is used, mk-table-checksum automatically sets
1334 "--throttle-method" to "slavelag" and discovers every slave and
1335 throttles checksumming if any slave lags more than "--max-lag".
1336 Specify "-throttle-method none" to disable this behavior
1337 completely, or specify "--check-slave-lag" and mk-table-checksum
1338 will only check that slave.
1339 See also "--recurse" and "--recursion-method".
1341 --trim
1343 Trim "VARCHAR" columns (helps when comparing 4.1 to >= 5.0).
1344 This option adds a "TRIM()" to "VARCHAR" columns in "BIT_XOR" and
1346 "ACCUM" modes.
1347 This is useful when you don't care about the trailing space
1349 differences between MySQL versions which vary in their handling of
1350 trailing spaces. MySQL 5.0 and later all retain trailing spaces in
1351 "VARCHAR", while previous versions would remove them.
1352 --unchunkable-tables
1354 group: Safety
1355 Checksum tables that cannot be chunked when "--chunk-size" is
1357 specified.
1358 By default mk-table-checksum will not checksum a table that cannot
1360 be chunked when "--chunk-size" is specified because this might
1361 result in a huge, non-chunkable table being checksummed in one
1362 huge, memory-intensive chunk.
1363 Specifying this option allows checksumming tables that cannot be
1365 chunked. Be careful when using this option! Make sure any non-
1366 chunkable tables are not so large that they will cause the tool to
1367 consume too much memory or CPU.
1368 See also "--chunk-size-limit".
1370 --[no]use-index
1372 default: yes
1373 Add FORCE INDEX hints to SQL statements.
1375 By default "mk-table-checksum" adds an index hint ("FORCE INDEX"
1377 for MySQL v4.0.9 and newer, "USE INDEX" for older MySQL versions)
1378 to each SQL statement to coerce MySQL into using the
1379 "--chunk-index" (whether the index is specified by the option or
1380 auto-detected). Specifying "--no-use-index" causes
1381 "mk-table-checksum" to omit index hints.
1382 --user
1384 short form: -u; type: string; group: Connection
1385 User for login if not current user.
1387 --[no]verify
1389 default: yes
1390 Verify checksum compatibility across servers.
1392 This option runs a trivial checksum on all servers to ensure they
1394 have compatible CONCAT_WS() and cryptographic hash functions.
1395 Versions of MySQL before 4.0.14 will skip empty strings and NULLs
1397 in CONCAT_WS, and others will only skip NULLs. The two kinds of
1398 behavior will produce different results if you have any columns
1399 containing the empty string in your table. If you know you don't
1400 (for instance, all columns are integers), you can safely disable
1401 this check and you will get a reliable checksum even on servers
1402 with different behavior.
1403 --version
1405 group: Help
1406 Show version and exit.
1408 --wait
1410 short form: -w; type: time
1411 Wait this long for slaves to catch up to their master (implies
1413 "--lock" "--slave-lag").
1414 Note: the best way to verify that a slave is in sync with its
1416 master is to use "--replicate" instead. The "--wait" option is
1417 really only useful if you're trying to compare masters and slaves
1418 without using "--replicate", which is possible but complex and less
1419 efficient in some ways.
1420 This option helps you get a consistent checksum across a master
1422 server and its slaves. It combines locking and waiting to
1423 accomplish this. First it locks the table on the master (the first
1424 server on the command line). Then it finds the master's binlog
1425 position. Checksums on slaves will be deferred until they reach
1426 the same binlog position.
1427 The argument to the option is the number of seconds to wait for the
1429 slaves to catch up to the master. It is actually the argument to
1430 MASTER_POS_WAIT(). If the slaves don't catch up to the master
1431 within this time, they will unblock and go ahead with the checksum.
1432 You can tell whether this happened by examining the STAT column in
1433 the output, which is the return value of MASTER_POS_WAIT().
1434 --where
1436 type: string
1437 Do only rows matching this "WHERE" clause (disallows "--algorithm"
1439 CHECKSUM).
1440 You can use this option to limit the checksum to only part of the
1442 table. This is particularly useful if you have append-only tables
1443 and don't want to constantly re-check all rows; you could run a
1444 daily job to just check yesterday's rows, for instance.
1445 This option is much like the -w option to mysqldump. Do not
1447 specify the WHERE keyword. You may need to quote the value. Here
1448 is an example:
1449 mk-table-checksum --where "foo=bar"
1451 --[no]zero-chunk
1453 default: yes
1454 Add a chunk for rows with zero or zero-equivalent values. The only
1456 has an effect when "--chunk-size" is specified. The purpose of the
1457 zero chunk is to capture a potentially large number of zero values
1458 that would imbalance the size of the first chunk. For example, if
1459 a lot of negative numbers were inserted into an unsigned integer
1460 column causing them to be stored as zeros, then these zero values
1461 are captured by the zero chunk instead of the first chunk and all
1462 its non-zero values.
1463
1465 These DSN options are used to create a DSN. Each option is given like
1466 "option=value". The options are case-sensitive, so P and p are not the
1467 same option. There cannot be whitespace before or after the "=" and if
1468 the value contains whitespace it must be quoted. DSN options are
1469 comma-separated. See the maatkit manpage for full details.
1470 · A
1472 dsn: charset; copy: yes
1474 Default character set.
1476 · D
1478 dsn: database; copy: yes
1480 Default database.
1482 · F
1484 dsn: mysql_read_default_file; copy: yes
1486 Only read default options from the given file
1488 · h
1490 dsn: host; copy: yes
1492 Connect to host.
1494 · p
1496 dsn: password; copy: yes
1498 Password to use when connecting.
1500 · P
1502 dsn: port; copy: yes
1504 Port number to use for connection.
1506 · S
1508 dsn: mysql_socket; copy: yes
1510 Socket file to use for connection.
1512 · u
1514 dsn: user; copy: yes
1516 User for login if not current user.
1518
1520 You can download Maatkit from Google Code at
1521 <http://code.google.com/p/maatkit/>, or you can get any of the tools
1522 easily with a command like the following:
1523 wget http://www.maatkit.org/get/toolname
1525 or
1526 wget http://www.maatkit.org/trunk/toolname
1527 Where "toolname" can be replaced with the name (or fragment of a name)
1529 of any of the Maatkit tools. Once downloaded, they're ready to run; no
1530 installation is needed. The first URL gets the latest released version
1531 of the tool, and the second gets the latest trunk code from Subversion.
1532
1534 The environment variable "MKDEBUG" enables verbose debugging output in
1535 all of the Maatkit tools:
1536 MKDEBUG=1 mk-....
1538
1540 You need Perl, DBI, DBD::mysql, and some core packages that ought to be
1541 installed in any reasonably new version of Perl.
1542
1544 For a list of known bugs see
1545 <http://www.maatkit.org/bugs/mk-table-checksum>.
1546 Please use Google Code Issues and Groups to report bugs or request
1548 support: <http://code.google.com/p/maatkit/>. You can also join
1549 #maatkit on Freenode to discuss Maatkit.
1550 Please include the complete command-line used to reproduce the problem
1552 you are seeing, the version of all MySQL servers involved, the complete
1553 output of the tool when run with "--version", and if possible,
1554 debugging output produced by running with the "MKDEBUG=1" environment
1555 variable.
1556
1558 This program is copyright 2007-2011 Baron Schwartz. Feedback and
1559 improvements are welcome.
1560 THIS PROGRAM IS PROVIDED "AS IS" AND WITHOUT ANY EXPRESS OR IMPLIED
1562 WARRANTIES, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTIES OF
1563 MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
1564 This program is free software; you can redistribute it and/or modify it
1566 under the terms of the GNU General Public License as published by the
1567 Free Software Foundation, version 2; OR the Perl Artistic License. On
1568 UNIX and similar systems, you can issue `man perlgpl' or `man
1569 perlartistic' to read these licenses.
1570 You should have received a copy of the GNU General Public License along
1572 with this program; if not, write to the Free Software Foundation, Inc.,
1573 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA.
1574
1576 See also mk-checksum-filter and mk-table-sync.
1577
1579 Baron "Xaprb" Schwartz
1580
1582 This tool is part of Maatkit, a toolkit for power users of MySQL.
1583 Maatkit was created by Baron Schwartz; Baron and Daniel Nichter are the
1584 primary code contributors. Both are employed by Percona. Financial
1585 support for Maatkit development is primarily provided by Percona and
1586 its clients.
1587
1589 This is an incomplete list. My apologies for omissions or
1590 misspellings.
1591 Claus Jeppesen, Francois Saint-Jacques, Giuseppe Maxia, Heikki Tuuri,
1593 James Briggs, Martin Friebe, Sergey Zhuravlev,
1594
1596 This manual page documents Ver 1.2.23 Distrib 7540 $Revision: 7527 $.
1597perl v5.28.0 2011-06-08 MK-TABLE-CHECKSUM(1)