Daniel Nichter
217a45df6f
Conditionalize top 5 slowest test files, saving 25%/5m * 15 configs = 1hr.
2012-08-12 10:16:52 -06:00
Daniel Nichter
a9eabffab9
Fix and stablize pt-slave-delay tests.
2012-08-11 13:52:41 -06:00
Daniel Nichter
3822bb7977
Hack a fix for MasterSlave in tests that use it.
2012-08-02 15:34:39 -06:00
Brian Fraser
ab8d26df4b
Merged fix-883098-extra-processlist-fields
2012-08-02 15:12:59 -03:00
Daniel Nichter
c80123f485
Merge fix-953141-recursion-method-array
2012-08-02 12:01:34 -06:00
Daniel Nichter
e6d31a666a
Rewrite how MasterSlave handles --recursion-methods.
2012-08-02 10:57:36 -06:00
Brian Fraser
e29aa514e3
t/lib/MasterSlave.t: A call to ->recurse_to_slaves() still used a string instead of an arrayref
2012-08-02 12:38:09 -03:00
Brian Fraser fraserb@gmail.com
654ebacf1c
Fix for 883098: Extra processlist fields causing errors
2012-08-01 18:43:03 -03:00
Daniel Nichter
cc3826f665
Fix more test output broken by percona_test.load_data. Tweak ptc/t/privs.t so it passes on my box. Fix new start/stop scripts: don't check ibdata1 because SKIP_INNODB may be used. Make PerconaTest::output() cleaner.
2012-08-01 14:58:52 -06:00
Brian Fraser fraserb@gmail.com
d99c7a6041
Update all modules that use --recursion-method, additionally changing it to an array with default processlist,hosts and checking the array for validity
2012-08-01 17:43:09 -03:00
Brian Fraser fraserb@gmail.com
a25b9eab40
MasterSlave: Rework how the recursion methods are checked & resolved
...
In part by removing the OptionParser usage out from get_slaves
and recurse_to_slaves and making them expect arrayrefs, thus
forcing our callers to deal with that, and in part by
splitting out the method-checking to MasterSlave::check_recursion_method
and the resolving (originally in find_slave_hosts) into
_resolve_recursion_methods.
2012-08-01 17:39:12 -03:00
Brian Fraser fraserb@gmail.com
ad46ec779e
Merged fix-889739-pt-cd-quoted-strings and added the new sample files
2012-08-01 15:02:28 -03:00
Brian Fraser fraserb@gmail.com
dd14121c28
Merged fix-903510-pt-tcp-model-empty-requests-file
2012-08-01 14:56:10 -03:00
Brian Fraser fraserb@gmail.com
700e90d2db
Merged fix-821696-pt-va-snippet-delimiters
2012-08-01 14:53:40 -03:00
Brian Fraser fraserb@gmail.com
66d6ed75c0
Merged fix-844880-variableadvisor-bin_log-checks
2012-08-01 14:47:50 -03:00
Brian Fraser fraserb@gmail.com
90c7f6c9b4
Merged fix-1022628-standarize-pt-osc-quiet
2012-08-01 13:49:51 -03:00
Brian Fraser fraserb@gmail.com
9596574e85
Merged fix-950294-ptc-if-not-exists
2012-08-01 13:39:32 -03:00
Daniel Nichter
69a9562e9e
Fix t/lib/samples/SchemaIterator/all-dbs-tbls* that I broke with percona_test.load_data from skip-load-data-file branch.
2012-08-01 10:04:55 -06:00
Daniel Nichter
c769f4df11
Update ComopareResults.t with new $can_load_data.
2012-07-31 16:34:38 -06:00
Daniel Nichter
570b26de20
Remove LOAD DATA check from pt-archiver and pt-upgrade. Make test-env check if LOAD DATA works and PerconaTest export . Rewrite how pt-archiver/t/bulk_insert.t conditionalizes on LOAD DATA working.
2012-07-31 15:37:02 -06:00
Brian Fraser fraserb@gmail.com
b5191a310a
Fixes as per Daniel's review; Removed the --quiet usage inpt-osc's nibble_is_safe, as it was ptc cruft
2012-07-31 13:33:30 -03:00
Brian Fraser fraserb@gmail.com
a94da84b27
Changes from Daniel's review.
...
Some wording changes, also fatalized two warnings, and some test
changes.
2012-07-30 19:10:45 -03:00
Brian Fraser fraserb@gmail.com
62c74e1e76
Previous commit left an unfinished comment
2012-07-30 19:08:37 -03:00
Brian Fraser fraserb@gmail.com
1de45b70be
pt-osc: Change --quiet back and remove the quietness checks
2012-07-30 18:35:24 -03:00
Brian Fraser fraserb@gmail.com
ba6253545d
Merged fix-821703-processlist-null-state
2012-07-30 16:48:15 -03:00
Brian Fraser fraserb@gmail.com
b93856a37d
Fix for 903510: pt-tcp-model throws an error in --type=requests mode on empty file
2012-07-30 11:52:53 -03:00
Brian Fraser fraserb@gmail.com
4c5678876d
Fix for bug 889739: pt-config-diff doesn't diff quoted strings properly
2012-07-30 11:50:42 -03:00
Brian Fraser fraserb@gmail.com
2192eb9e4e
pt-va: Delimit sentences by . or ?
2012-07-30 11:47:59 -03:00
Brian Fraser fraserb@gmail.com
43708d7b65
VariableAdvisorRules: Check that log_bin is ON instead of doing log_bin ? 1 : 0
2012-07-30 11:44:28 -03:00
Brian Fraser fraserb@gmail.com
05f923545c
pt-osc: Make --quiet behave like in the other tools
2012-07-30 11:41:09 -03:00
Brian Fraser fraserb@gmail.com
2f0812fb33
Merged ubuntu-12-64-issues
2012-07-30 11:30:05 -03:00
Daniel Nichter
41c8aa001d
Failing test for --database bug 1028614.
2012-07-28 14:17:28 -06:00
Brian Fraser
192f38498c
Removed the charset munging and made a note in the pt-archiver docs that the only legal charsets are those known by MySQL
2012-07-27 16:09:32 -03:00
Daniel Nichter
0fc84ff7c0
Switch bugs.t to use done_testing.
2012-07-27 11:52:24 -06:00
Daniel Nichter
203b0a9cb6
Report DIFFS as the true number of unique chunks that differ across all slaves.
2012-07-27 11:30:33 -06:00
Daniel Nichter
1875c92674
Partial bug fix: report max diffs, not just last slave's diffs. Still needs fixing to report count of all unique chunk# diffs.
2012-07-27 11:16:41 -06:00
Daniel Nichter
a07dc3c146
Merge fix-pt-osc-empty-table-bug-1020997
2012-07-26 16:49:40 -06:00
Daniel Nichter
2290827eee
Stop nibbling if chunking but not first lower boundary because table is empty. Print test debug via diag().
2012-07-26 16:44:24 -06:00
Daniel Nichter
b59040eb83
Test that pt-archiver DSN b=1 causes SET SQL_LOG_BIN=0.
2012-07-26 14:24:07 -06:00
Brian Fraser
776798b3b9
Make pt-table-checksum always create the database/table it needs
2012-07-25 17:10:30 -03:00
Brian Fraser
0d26097f31
t/pt-table-sync/wait.t: use wait_until() instead of sleep()
2012-07-23 13:48:05 -03:00
Brian Fraser
a948014991
t/lib/Daemon.t: Tweak the tests to hopefully avoid white smoke
2012-07-23 12:28:55 -03:00
Brian Fraser
a1f1168ab4
t/lib/MasterSlave.t: Check that recurse_to_slaves gets all three slaves, or give a diagnostic
2012-07-23 12:28:25 -03:00
Brian Fraser
b6e4202cfa
t/pt-fifo-split/pt-fifo-split.t: Modernize the tests to use wait_for_files and give better diagnostics
2012-07-23 12:26:09 -03:00
Brian Fraser
a195c09bc3
t/pt-ioprofile/pt-ioprofile.t: Mark the --run-time test as TODO to avoid white smoke
2012-07-23 12:17:37 -03:00
Brian Fraser
68b8eddc5a
t/lib/bash/collect.sh: Wait for the df file before starting
2012-07-23 02:43:15 -03:00
Brian Fraser
5ac234d1b2
t/pt-archiver/check_slave_lag.t: Mark a timing-related test as TODO to avoid white smoke
2012-07-23 02:06:50 -03:00
Brian Fraser
409f40b240
t/pt-deadlock-logger/standard_options.t: Modernize the tests to stop relying on sleep(), modify an ancient test that was looking for mt-deadlock-logger
2012-07-23 01:52:41 -03:00
Brian Fraser
61a65fcdbd
DSNParser.pm: Work around MySQL not resolving encodings sanely.
...
For some reason, MySQL won't resolve iso-8859-1 or latin-1 as latin1,
and will die if you try to use through in, e.g., SET NAMES.
This commit makes DSNParser resolve those cases on its own
before calling SET NAMES.
2012-07-23 01:11:49 -03:00
Brian Fraser
02c6421a25
Cxn.pm: Set NAME_lc by default
2012-07-22 23:37:39 -03:00
