summaryrefslogtreecommitdiff
path: root/mysql-test/t/partition_innodb.test
diff options
context:
space:
mode:
authorDmitry Lenev <Dmitry.Lenev@oracle.com>2010-11-19 10:26:09 +0300
committerDmitry Lenev <Dmitry.Lenev@oracle.com>2010-11-19 10:26:09 +0300
commit602a22259e5e6e7c6f8f43ec31b81368f04182a3 (patch)
tree4e0486b07de2f5676e27e519f22a1d83606f94c8 /mysql-test/t/partition_innodb.test
parent7771250d2bcddd5a00eb8ce854527eafb0e833a3 (diff)
downloadmariadb-git-602a22259e5e6e7c6f8f43ec31b81368f04182a3.tar.gz
Fix for bug #57985 "ONLINE/FAST ALTER PARTITION can fail and
leave the table unusable". Failing ALTER statement on partitioned table could have left this table in an unusable state. This has happened in cases when ALTER was executed using "fast" algorithm, which doesn't involve copying of data between old and new versions of table, and the resulting new table was incompatible with partitioning function in some way. The problem stems from the fact that discrepancies between new table definition and partitioning function are discovered only when the table is opened. In case of "fast" algorithm this has happened too late during ALTER's execution, at the moment when all changes were already done and couldn't have been reverted. In the cases when "slow" algorithm, which copies data, is used such discrepancies are detected at the moment new table definition is opened implicitly when new version of table is created in storage engine. As result ALTER is aborted before any changes to table were done. This fix tries to address this issue by ensuring that "fast" algorithm behaves similarly to "slow" algorithm and checks compatibility between new definition and partitioning function by trying to open new definition after .FRM file for it has been created. Long term we probably should implement some way to check compatibility between partitioning function and new table definition which won't involve opening it, as this should allow much cleaner fix for this problem. mysql-test/r/partition_innodb.result: Added test for bug #57985 "ONLINE/FAST ALTER PARTITION can fail and leave the table unusable". mysql-test/t/partition_innodb.test: Added test for bug #57985 "ONLINE/FAST ALTER PARTITION can fail and leave the table unusable". sql/sql_table.cc: Ensure that in cases when .FRM for partitioned table is created without creating table in storage engine (e.g. during "fast" ALTER TABLE) we still open table definition. This allows to check that definition of created table/.FRM is compatible with its partitioning function.
Diffstat (limited to 'mysql-test/t/partition_innodb.test')
-rw-r--r--mysql-test/t/partition_innodb.test21
1 files changed, 21 insertions, 0 deletions
diff --git a/mysql-test/t/partition_innodb.test b/mysql-test/t/partition_innodb.test
index 3e9ac2ce2b5..ddc2812f617 100644
--- a/mysql-test/t/partition_innodb.test
+++ b/mysql-test/t/partition_innodb.test
@@ -569,3 +569,24 @@ SET SESSION sql_mode = 'NO_ZERO_DATE';
OPTIMIZE TABLE t1;
SET SESSION sql_mode = @old_mode;
DROP TABLE t1;
+
+
+--echo #
+--echo # Bug#57985 "ONLINE/FAST ALTER PARTITION can fail and leave the
+--echo # table unusable".
+--echo #
+--disable_warnings
+DROP TABLE IF EXISTS t1;
+--enable_warnings
+CREATE TABLE t1 (a bigint not null, b int not null, PRIMARY KEY (a))
+ ENGINE = InnoDB PARTITION BY KEY(a) PARTITIONS 2;
+INSERT INTO t1 values (0,1), (1,2);
+--echo # The below ALTER should fail. It should leave the
+--echo # table in its original, non-corrupted, usable state.
+--error ER_UNIQUE_KEY_NEED_ALL_FIELDS_IN_PF
+ALTER TABLE t1 ADD UNIQUE KEY (b);
+--echo # The below statements should succeed, as ALTER should
+--echo # have left table intact.
+SHOW CREATE TABLE t1;
+SELECT * FROM t1;
+DROP TABLE t1;