Fix tests for pt-table-checksum

$ prove -vw --trap --timer t/pt-table-checksum/
[20:14:42] t/pt-table-checksum/basics.t .................. 
1..40
not ok 1 - Default checksum

#   Failed test 'Default checksum'
#   at t/pt-table-checksum/basics.t line 75.
ok 2 - Between 25 and 60 chunks
not ok 3 - Static chunk size (--chunk-time 0)

#   Failed test 'Static chunk size (--chunk-time 0)'
#   at t/pt-table-checksum/basics.t line 101.
ok 4 - Between 75 and 90 chunks on master
ok 5 - ... same number of chunks on slave
Checking if all tables can be checksummed ...
Starting checksum ...
            TS ERRORS  DIFFS     ROWS  DIFF_ROWS  CHUNKS SKIPPED    TIME TABLE
12-20T20:15:39      0      1      600          0     602       0  22.358 sakila.city
ok 6 - --replicate-check on by default, detects diff
Checking if all tables can be checksummed ...
Starting checksum ...
ok 7 - --no-replicate-check, no diff detected
Checking if all tables can be checksummed ...
Starting checksum ...
Checking if all tables can be checksummed ...
Starting checksum ...
ok 8 - --emptry-replicate-table on by default
Checking if all tables can be checksummed ...
Starting checksum ...
ok 9 - --no-recheck (just --replicate-check)
ok 10 - Detects infinite loop
ok 11 - Upper boundary same as next lower boundary
ok 12 - Skipped oversize chunks
ok 13 - Oversize chunks are not errors
ok 14 - Skipped chunk 2
ok 15 - Skipped 1 chunk (bug 1011738)
ok 16 - Warns about skipping large slave table
ok 17 - Does not checksum large slave table on master
ok 18 - Does not checksum large slave table on slave
ok 19 - Non-zero exit status
ok 20 - 0 skipped
ok 21 - 0 errors
ok 22 - 52 rows checksummed
ok 23 - --chunk-size-limit=0 does not disable \#-of-rows checks on slaves
ok 24 - Warns when no slave are found (bug 1087804)
ok 25 - Exit status 8 when no slaves are found (bug 1087804)
ok 26 - --where in checksum chunk query
ok 27 - --where in past lower chunk query
ok 28 - --where in past upper chunk query
ok 29 - --where in next chunk boundary query
ok 30 - --where for first chunk
ok 31 - --where for last chunk
ok 32 - --where for lower oob chunk
ok 33 - --where for upper oob chunk
ok 34 - Bug 932442: 0 exit
ok 35 - Bug 932442: 0 errors
ok 36 - Bug 821675 (dot): queries
ok 37 - Bug 821675 (dot): 0 exit
ok 38 - Bug 821675 (dot): 0 errors
Checking if all tables can be checksummed ...
Starting checksum ...
ok 39 - sql_mode ONLY_FULL_GROUP_BY is overidden
ok 40 - Sandbox servers
# Looks like you failed 2 tests of 40.
Dubious, test returned 2 (wstat 512, 0x200)
Failed 2/40 subtests 
[20:16:32] t/pt-table-checksum/bugs.t .................... 
ok 1 - Bug 995274 (undef array): zero exit status
ok 2 - Bug 995274 (undef array): checksummed rows
ok 3 - Bug 987393 (empty table): zero exit status
ok 4 - Bug 987393 (empty table): no errors
ok 5 - Bug 987393 (empty table): checksums
ok 6 - Bug 987393 (Perl 5.8 scoping): no errors
not ok 7 - Bug 987393 (Perl 5.8 scoping): checksummed table

#   Failed test 'Bug 987393 (Perl 5.8 scoping): checksummed table'
#   at t/pt-table-checksum/bugs.t line 2281.
#          got: '0'
#     expected: '109'
ok 8 - Bug 1030031 (wrong DIFFS): 2 diffs
ok 9 - Bug 1030031 (wrong DIFFS): 3 diffs
ok 10 - Bug 1074179: ignore-tables-regex works with --replicate-check-only
ok 11 - Bug 1016131: ptc should skip tables where all columns are excluded
ok 12 - --skip-check-slave-lag
ok 13 - Bugs 1075638 and 1050737: No division by zero error when nibble_time is zero
ok 14 - Bug 1075638 and 1050737: ...And we get the correct number of diffs
# Creating default databases ...
# Starting MySQL test server on port 12348... return 0
# OK (3s)
# Creating default databases ...
# Starting MySQL test server on port 12349... return 0
# OK (2s)
ok 15 - Bug 938068: doesn't warn if binlog_format=row or mixed on slaves
ok 16 - ...and warns for both level 1 and level 2 slaves
# Stopping MySQL test server on port 12349... OK (4s)
# Stopping MySQL test server on port 12348... OK (2s)
ok 17 - Bug 1210537: no crash with --recursion-method cluster
ok 18 - Bug 1210537: tool ran
ok 19 - Bug 1388870 - No false positive reported when system_tz differ on slave
ok 20 - Sandbox servers
1..20
ok 6 - Bug 987393 (Perl 5.8 scoping): no errors
not ok 7 - Bug 987393 (Perl 5.8 scoping): checksummed table

#   Failed test 'Bug 987393 (Perl 5.8 scoping): checksummed table'
#   at t/pt-table-checksum/bugs.t line 2281.
#          got: '0'
#     expected: '109'
DBD::mysql::db do failed: Lost connection to MySQL server during query [for Statement "UPDATE test.t SET c='' WHERE id=5"] at t/pt-table-checksum/bugs.t line 2295.
# Tests were run but no plan was declared and done_testing() was not seen.
# Looks like your test exited with 255 just after 7.
Dubious, test returned 255 (wstat 65280, 0xff00)
Failed 1/20 subtests 
[20:18:52] t/pt-table-checksum/char_chunking.t ........... 
1..6
ok 1 - Char chunk ascii, explain
ok 2 - Char chunk ascii, chunk size 20
ok 3 - First boundaries
ok 4 - Lower oob boundary
ok 5 - Upper oob boundary
ok 6 - Sandbox servers
ok     7206 ms ( 0.01 usr  0.00 sys +  0.90 cusr  0.49 csys =  1.40 CPU)
[20:18:59] t/pt-table-checksum/chunk_index.t ............. 
1..14
ok 1 - Chooses chunk index by default
ok 2 - Chooses chunk index if --chunk-index doesn't exist
ok 3 - Use --chunk-index
ok 4 - Chunks on left-most --chunk-index column
ok 5 - Explicit --chunk-index overrides MySQL's index for --where
ok 6 - Zero exit status (bug 925855)
ok 7 - 0 skipped (bug 925855)
ok 8 - 0 errors (bug 925855)
ok 9 - 14 rows checksummed (bug 925855)
ok 10 - Smarter chunk index selection (bug 978432)
ok 11 - --chunk-index-columns
ok 12 - --chunk-index-columns > number of index columns
ok 13 - Initial key_len reflects --chunk-index-columns
ok 14 - Sandbox servers
ok     5965 ms ( 0.00 usr  0.00 sys +  0.98 cusr  0.46 csys =  1.44 CPU)
[20:19:05] t/pt-table-checksum/chunk_size.t .............. 
1..8
Checking if all tables can be checksummed ...
Starting checksum ...
ok 1 - First chunk is default size
ok 2 - 2nd chunk lower boundary
ok 3 - 2nd chunk is larger
Checking if all tables can be checksummed ...
Starting checksum ...
ok 4 - Explicit --chunk-size disables auto chunk sizing
ok 5 # skip Too slow
DBD::mysql::db do failed: LOAD DATA LOCAL INFILE file request rejected due to restrictions on access. [for Statement "LOAD DATA LOCAL INFILE '/home/sveta/src/percona-toolkit/t/pt-table-checksum/samples/600cities.data' INTO TABLE test.t"] at t/pt-table-checksum/chunk_size.t line 114.
# Looks like your test exited with 255 just after 5.
Dubious, test returned 255 (wstat 65280, 0xff00)
Failed 3/8 subtests 
	(less 1 skipped subtest: 4 okay)
[20:19:08] t/pt-table-checksum/create_replicate_table.t .. 
1..5
Unknown option: no-create-replicate-table
Usage: pt-table-checksum [OPTIONS] [DSN]

Errors in command-line arguments:
  * Error parsing options

pt-table-checksum performs an online replication consistency check by executing
checksum queries on the master, which produces different results on replicas
that are inconsistent with the master.  The optional DSN specifies the master
host.  The tool's L<"EXIT STATUS"> is non-zero if any differences are found, or
if any warnings or errors occur.  For more details, please use the --help
option, or try 'perldoc t/pt-table-checksum/create_replicate_table.t' for
complete documentation.
# Looks like your test exited with 1 before it could output anything.
Dubious, test returned 1 (wstat 256, 0x100)
Failed 5/5 subtests 
[20:19:09] t/pt-table-checksum/error_handling.t .......... 
1..0 # SKIP TODO master master sandbox is failing with MySQL 8.0+. FIX ME !!!!
skipped: TODO master master sandbox is failing with MySQL 8.0+. FIX ME !!!!
[20:19:09] t/pt-table-checksum/filters.t ................. 
1..9
ok 1 - --databases sakila
ok 2 - --tables actor
ok 3 - --tables sakila.actor
ok 4 - --tables actor,film
ok 5 - --tables sakila.actor,mysql.user
ok 6 - -d sakila --engines MyISAM
ok 7 - -d sakila --engines myisam
ok 8 - --databases sakila --ignore-tables sakila.actor,sakila.address,category,city,payment
ok 9 - Sandbox servers
ok     1572 ms ( 0.01 usr  0.00 sys +  0.50 cusr  0.24 csys =  0.75 CPU)
[20:19:10] t/pt-table-checksum/float_precision.t ......... 
1..7
ok 1 - Got output
ok 2 - No --float-precision, no rounding
ok 3 - Got output
ok 4 - Column a is rounded
ok 5 - Column b is rounded
ok 6 - Column b is not rounded inside ISNULL
ok 7 - Sandbox servers
ok     3126 ms ( 0.00 usr  0.00 sys +  0.36 cusr  0.17 csys =  0.53 CPU)
[20:19:14] t/pt-table-checksum/fnv_64.t .................. 
1..0 # SKIP No FNV_64 UDF lib
skipped: No FNV_64 UDF lib
[20:19:14] t/pt-table-checksum/ignore_columns.t .......... 
1..4
ok 1 - Diff when column not ignored
ok 2 - No diff when column ignored
ok 3 - Ignored column is not in checksum query
ok 4 - Sandbox servers
ok     3125 ms ( 0.00 usr  0.00 sys +  0.39 cusr  0.20 csys =  0.59 CPU)
[20:19:17] t/pt-table-checksum/issue_1485195.t ........... 
1..2
not ok 1 - Large BLOB/TEXT/BINARY Checksum

#   Failed test 'Large BLOB/TEXT/BINARY Checksum'
#   at t/pt-table-checksum/issue_1485195.t line 43.
#          got: '31'
#     expected: '29'
ok 2 - Sandbox servers
# Looks like you failed 1 test of 2.
Dubious, test returned 1 (wstat 256, 0x100)
Failed 1/2 subtests 
[20:19:20] t/pt-table-checksum/issue_1592608.t ........... 
1..2
ok 1 - Large BLOB/TEXT/BINARY Checksum
ok 2 - Sandbox servers
ok     4336 ms ( 0.00 usr  0.00 sys +  0.37 cusr  0.16 csys =  0.53 CPU)
[20:19:25] t/pt-table-checksum/issue_1651002.t ........... 
1..2
ok 1 - Large BLOB/TEXT/BINARY Checksum
# Stopping the sandbox to leave a clean sandbox for the next test file
ok 2 - Sandbox servers
ok     3809 ms ( 0.00 usr  0.00 sys +  0.50 cusr  0.30 csys =  0.80 CPU)
[20:19:28] t/pt-table-checksum/issue_388.t ............... 
1..3
ok 1 - No error (issue 388)
ok 2 - Checksums the table (issue 388)
ok 3 - Sandbox servers
ok     2313 ms ( 0.00 usr  0.00 sys +  0.40 cusr  0.13 csys =  0.53 CPU)
[20:19:31] t/pt-table-checksum/issue_47.t ................ 
1..3
Checking if all tables can be checksummed ...
Starting checksum ...
12-20T20:19:32 Skipping chunk 5 of test.issue_47 because MySQL chose no index  instead of the idxindex.
not ok 1 - No error nibbling very large int

#   Failed test 'No error nibbling very large int'
#   at t/pt-table-checksum/issue_47.t line 50.
#          got: '32'
#     expected: '0'
not ok 2 - Uses very large int as chunk boundary

#   Failed test 'Uses very large int as chunk boundary'
#   at t/pt-table-checksum/issue_47.t line 57.
#     Structures begin differing at:
#          $got->[4] = Does not exist
#     $expected->[4] = ARRAY(0x55b2d2826948)
ok 3 - Sandbox servers
# Looks like you failed 2 tests of 3.
Dubious, test returned 2 (wstat 512, 0x200)
Failed 2/3 subtests 
[20:19:33] t/pt-table-checksum/issue_602.t ............... 
1..2
ok 1 - Checksums table despite invalid datetime
ok 2 - Sandbox servers
ok     3027 ms ( 0.00 usr  0.00 sys +  0.34 cusr  0.17 csys =  0.51 CPU)
[20:19:36] t/pt-table-checksum/option_sanity.t ........... Can't exec "/home/sveta/src/percona-toolkit/bin/pt-table-checksum": Permission denied at t/pt-table-checksum/option_sanity.t line 28.

not ok 1 - Default --check-replication-filters=TRUE

#   Failed test 'Default --check-replication-filters=TRUE'
#   at t/pt-table-checksum/option_sanity.t line 30.
#                   ''
#     doesn't match '(?^m:^  --check-replication-filters\s+TRUE$)'
not ok 2 - Default --create-replicate-table=TRUE

#   Failed test 'Default --create-replicate-table=TRUE'
#   at t/pt-table-checksum/option_sanity.t line 36.
#                   ''
#     doesn't match '(?^m:^  --create-replicate-table\s+TRUE$)'
not ok 3 - Default --empty-replicate-table=TRUE

#   Failed test 'Default --empty-replicate-table=TRUE'
#   at t/pt-table-checksum/option_sanity.t line 42.
#                   ''
#     doesn't match '(?^m:^  --empty-replicate-table\s+TRUE$)'
not ok 4 - Default --explain=0

#   Failed test 'Default --explain=0'
#   at t/pt-table-checksum/option_sanity.t line 48.
#                   ''
#     doesn't match '(?^m:^  --explain\s+0$)'
not ok 5 - Default --host=localhost

#   Failed test 'Default --host=localhost'
#   at t/pt-table-checksum/option_sanity.t line 54.
#                   ''
#     doesn't match '(?^m:^  --host\s+localhost$)'
not ok 6 - Default --max-lag=1

#   Failed test 'Default --max-lag=1'
#   at t/pt-table-checksum/option_sanity.t line 66.
#                   ''
#     doesn't match '(?^m:^  --max-lag\s+1$)'
not ok 7 - Default --quiet=0

#   Failed test 'Default --quiet=0'
#   at t/pt-table-checksum/option_sanity.t line 72.
#                   ''
#     doesn't match '(?^m:^  --quiet\s+0$)'
not ok 8 - Default --replicate-check-only=FALSE

#   Failed test 'Default --replicate-check-only=FALSE'
#   at t/pt-table-checksum/option_sanity.t line 78.
#                   ''
#     doesn't match '(?^m:^  --replicate-check-only\s+FALSE$)'
not ok 9 - Default --replicate=percona.checksums

#   Failed test 'Default --replicate=percona.checksums'
#   at t/pt-table-checksum/option_sanity.t line 84.
#                   ''
#     doesn't match '(?^m:^  --replicate\s+percona\.checksums$)'
not ok 10 - Default --replicate-check=TRUE

#   Failed test 'Default --replicate-check=TRUE'
#   at t/pt-table-checksum/option_sanity.t line 90.
#                   ''
#     doesn't match '(?^m:^  --replicate-check\s+TRUE$)'
not ok 11 - --recursion-method is an array

#   Failed test '--recursion-method is an array'
#   at t/pt-table-checksum/option_sanity.t line 96.
#                   ''
#     doesn't match '(?^m:^\s+--recursion-method=a)'
not ok 12 - Default --recursion-method is processlist,hosts

#   Failed test 'Default --recursion-method is processlist,hosts'
#   at t/pt-table-checksum/option_sanity.t line 102.
#                   ''
#     doesn't match '(?^m:^\s+--recursion-method\s+processlist,hosts)'
Can't exec "/home/sveta/src/percona-toolkit/bin/pt-table-checksum": Permission denied at t/pt-table-checksum/option_sanity.t line 111.
not ok 13 - --explain disables --empty-replicate-table

#   Failed test '--explain disables --empty-replicate-table'
#   at t/pt-table-checksum/option_sanity.t line 112.
#                   ''
#     doesn't match '(?^m:^  --empty-replicate-table\s+FALSE$)'
Can't exec "/home/sveta/src/percona-toolkit/bin/pt-table-checksum": Permission denied at t/pt-table-checksum/option_sanity.t line 118.
not ok 14 - --resume disables --empty-replicate-table

#   Failed test '--resume disables --empty-replicate-table'
#   at t/pt-table-checksum/option_sanity.t line 119.
#                   ''
#     doesn't match '(?^m:^  --empty-replicate-table\s+FALSE$)'
Can't exec "/home/sveta/src/percona-toolkit/bin/pt-table-checksum": Permission denied at t/pt-table-checksum/option_sanity.t line 125.
not ok 15 - --quiet disables --progress

#   Failed test '--quiet disables --progress'
#   at t/pt-table-checksum/option_sanity.t line 126.
#                   ''
#     doesn't match '(?^m:^  --progress\s+\(No value\)$)'
Can't exec "/home/sveta/src/percona-toolkit/bin/pt-table-checksum": Permission denied at t/pt-table-checksum/option_sanity.t line 132.
not ok 16 - --chunk-size sets --chunk-time=0

#   Failed test '--chunk-size sets --chunk-time=0'
#   at t/pt-table-checksum/option_sanity.t line 133.
#                   ''
#     doesn't match '(?^m:^  --chunk-time\s+0$)'
Can't exec "/home/sveta/src/percona-toolkit/bin/pt-table-checksum": Permission denied at t/pt-table-checksum/option_sanity.t line 142.
not ok 17 - Only one DSN allowed on the command line

#   Failed test 'Only one DSN allowed on the command line'
#   at t/pt-table-checksum/option_sanity.t line 143.
#                   ''
#     doesn't match '(?^:More than one host specified; only one allowed)'
Can't exec "/home/sveta/src/percona-toolkit/bin/pt-table-checksum": Permission denied at t/pt-table-checksum/option_sanity.t line 152.
not ok 18 - --replicate table must be database-qualified

#   Failed test '--replicate table must be database-qualified'
#   at t/pt-table-checksum/option_sanity.t line 153.
#                   ''
#     doesn't match '(?^:--replicate table must be database-qualified)'
Can't exec "/home/sveta/src/percona-toolkit/bin/pt-table-checksum": Permission denied at t/pt-table-checksum/option_sanity.t line 162.
not ok 19 - --chunk-size-limit must be >= 1 or 0

#   Failed test '--chunk-size-limit must be >= 1 or 0'
#   at t/pt-table-checksum/option_sanity.t line 163.
#                   ''
#     doesn't match '(?^:chunk-size-limit must be >= 1 or 0 to disable)'
Can't exec "/home/sveta/src/percona-toolkit/bin/pt-table-checksum": Permission denied at t/pt-table-checksum/option_sanity.t line 173.
not ok 20 - Validates --max-load

#   Failed test 'Validates --max-load'
#   at t/pt-table-checksum/option_sanity.t line 174.
#                   ''
#     doesn't match '(?^:Invalid --max-load)'
1..20
# Looks like you failed 20 tests of 20.
Dubious, test returned 20 (wstat 5120, 0x1400)
Failed 20/20 subtests 
[20:19:36] t/pt-table-checksum/plugin.t .................. 
ok 1 - Called all plugins on basic run
ok 2 - Called all plugins on replicate-check run
ok 3 - Sandbox servers
1..3
ok     2308 ms ( 0.00 usr  0.00 sys +  0.41 cusr  0.16 csys =  0.57 CPU)
[20:19:39] t/pt-table-checksum/privs.t ................... 
ok 1 - Error if percona db doesn't exist and user can't create it
# Tests were run but no plan was declared and done_testing() was not seen.
# Looks like your test exited with 1 just after 1.
Dubious, test returned 1 (wstat 256, 0x100)
All 1 subtests passed 
[20:19:39] t/pt-table-checksum/progress.t ................ 
1..5
ok 1 - --progress for slave lag
ok 2 - Checksumming continues after waiting for slave lag
ok 3 - No errors after waiting for slave lag
ok 4 - --progress for --replicate table (bug 1008778)
ERROR: Tables are different on master: mysql.db
ERROR: Tables are different on slave1: mysql.db
ERROR: Tables are different on slave2: mysql.db
not ok 5 - Sandbox servers

#   Failed test 'Sandbox servers'
#   at t/pt-table-checksum/progress.t line 146.
Bailout called.  Further testing stopped:  t/pt-table-checksum/progress.t broke the sandbox
Bail out!  t/pt-table-checksum/progress.t broke the sandbox
FAILED--Further testing stopped: t/pt-table-checksum/progress.t broke the sandbox