Bug#41098: Query Cache returns wrong result with concurrent insert
The problem is that select queries executed concurrently with a concurrent insert on a MyISAM table could be cached if the select started after the query cache invalidation but before the unlock of tables performed by the concurrent insert. This race could happen because the concurrent insert was failing to prevent cache of select queries happening at the same time. The solution is to add a 'uncacheable' status flag to signal that a concurrent insert is being performed on the table and that queries executing at the same time shouldn't cache the results.
This commit is contained in:
parent
123549e7e8
commit
a2e4449bfc
@ -22,3 +22,52 @@ Qcache_queries_in_cache 0
|
||||
set global query_cache_size= 0;
|
||||
use test;
|
||||
drop table t1;
|
||||
SET @old_concurrent_insert= @@GLOBAL.concurrent_insert;
|
||||
SET @old_query_cache_size= @@GLOBAL.query_cache_size;
|
||||
DROP TABLE IF EXISTS t1, t2;
|
||||
CREATE TABLE t1 (a INT);
|
||||
CREATE TABLE t2 (a INT);
|
||||
INSERT INTO t1 VALUES (1),(2),(3);
|
||||
SET GLOBAL concurrent_insert= 1;
|
||||
SET GLOBAL query_cache_size= 1024*512;
|
||||
SET GLOBAL query_cache_type= ON;
|
||||
# Switch to connection con1
|
||||
SET SESSION debug='+d,wait_after_query_cache_invalidate';
|
||||
# Send concurrent insert, will wait in the query cache table invalidate
|
||||
INSERT INTO t1 VALUES (4);
|
||||
# Switch to connection default
|
||||
# Wait for concurrent insert to reach the debug point
|
||||
# Switch to connection con2
|
||||
# Send SELECT that shouldn't be cached
|
||||
SELECT * FROM t1;
|
||||
a
|
||||
1
|
||||
2
|
||||
3
|
||||
# Switch to connection default
|
||||
# Notify the concurrent insert to proceed
|
||||
SELECT ID FROM INFORMATION_SCHEMA.PROCESSLIST
|
||||
WHERE STATE = 'wait_after_query_cache_invalidate' INTO @thread_id;
|
||||
KILL QUERY @thread_id;
|
||||
# Switch to connection con1
|
||||
# Gather insert result
|
||||
SHOW STATUS LIKE "Qcache_queries_in_cache";
|
||||
Variable_name Value
|
||||
Qcache_queries_in_cache 0
|
||||
# Test that it's cacheable
|
||||
SELECT * FROM t1;
|
||||
a
|
||||
1
|
||||
2
|
||||
3
|
||||
4
|
||||
SHOW STATUS LIKE "Qcache_queries_in_cache";
|
||||
Variable_name Value
|
||||
Qcache_queries_in_cache 1
|
||||
# Disconnect
|
||||
# Restore defaults
|
||||
RESET QUERY CACHE;
|
||||
DROP TABLE t1,t2;
|
||||
SET GLOBAL concurrent_insert= DEFAULT;
|
||||
SET GLOBAL query_cache_size= DEFAULT;
|
||||
SET GLOBAL query_cache_type= DEFAULT;
|
||||
|
@ -10,5 +10,4 @@
|
||||
#
|
||||
##############################################################################
|
||||
kill : Bug#37780 2008-12-03 HHunger need some changes to be robust enough for pushbuild.
|
||||
query_cache_28249 : Bug#41098 Query Cache returns wrong result with concurrent insert
|
||||
innodb_bug39438 : BUG#42383 2009-01-28 lsoares "This fails in embedded and on windows. Note that this test is not run on windows and on embedded in PB for main trees currently"
|
||||
|
@ -44,3 +44,71 @@ set global query_cache_size= 0;
|
||||
use test;
|
||||
drop table t1;
|
||||
|
||||
#
|
||||
# Bug#41098: Query Cache returns wrong result with concurrent insert
|
||||
#
|
||||
|
||||
SET @old_concurrent_insert= @@GLOBAL.concurrent_insert;
|
||||
SET @old_query_cache_size= @@GLOBAL.query_cache_size;
|
||||
|
||||
--disable_warnings
|
||||
DROP TABLE IF EXISTS t1, t2;
|
||||
--enable_warnings
|
||||
CREATE TABLE t1 (a INT);
|
||||
CREATE TABLE t2 (a INT);
|
||||
INSERT INTO t1 VALUES (1),(2),(3);
|
||||
|
||||
SET GLOBAL concurrent_insert= 1;
|
||||
SET GLOBAL query_cache_size= 1024*512;
|
||||
SET GLOBAL query_cache_type= ON;
|
||||
|
||||
connect(con1,localhost,root,,test,,);
|
||||
connect(con2,localhost,root,,test,,);
|
||||
|
||||
connection con1;
|
||||
--echo # Switch to connection con1
|
||||
SET SESSION debug='+d,wait_after_query_cache_invalidate';
|
||||
--echo # Send concurrent insert, will wait in the query cache table invalidate
|
||||
--send INSERT INTO t1 VALUES (4)
|
||||
|
||||
connection default;
|
||||
--echo # Switch to connection default
|
||||
--echo # Wait for concurrent insert to reach the debug point
|
||||
let $wait_condition=
|
||||
SELECT COUNT(*) = 1 FROM INFORMATION_SCHEMA.PROCESSLIST
|
||||
WHERE STATE = "wait_after_query_cache_invalidate" AND
|
||||
INFO = "INSERT INTO t1 VALUES (4)";
|
||||
--source include/wait_condition.inc
|
||||
|
||||
connection con2;
|
||||
--echo # Switch to connection con2
|
||||
--echo # Send SELECT that shouldn't be cached
|
||||
SELECT * FROM t1;
|
||||
|
||||
connection default;
|
||||
--echo # Switch to connection default
|
||||
--echo # Notify the concurrent insert to proceed
|
||||
SELECT ID FROM INFORMATION_SCHEMA.PROCESSLIST
|
||||
WHERE STATE = 'wait_after_query_cache_invalidate' INTO @thread_id;
|
||||
KILL QUERY @thread_id;
|
||||
|
||||
connection con1;
|
||||
--echo # Switch to connection con1
|
||||
--echo # Gather insert result
|
||||
--reap
|
||||
SHOW STATUS LIKE "Qcache_queries_in_cache";
|
||||
--echo # Test that it's cacheable
|
||||
SELECT * FROM t1;
|
||||
SHOW STATUS LIKE "Qcache_queries_in_cache";
|
||||
|
||||
--echo # Disconnect
|
||||
disconnect con1;
|
||||
disconnect con2;
|
||||
|
||||
connection default;
|
||||
--echo # Restore defaults
|
||||
RESET QUERY CACHE;
|
||||
DROP TABLE t1,t2;
|
||||
SET GLOBAL concurrent_insert= DEFAULT;
|
||||
SET GLOBAL query_cache_size= DEFAULT;
|
||||
SET GLOBAL query_cache_type= DEFAULT;
|
||||
|
@ -1518,6 +1518,9 @@ void Query_cache::invalidate(THD *thd, TABLE_LIST *tables_used,
|
||||
invalidate_table(thd, tables_used);
|
||||
}
|
||||
|
||||
DBUG_EXECUTE_IF("wait_after_query_cache_invalidate",
|
||||
debug_wait_for_kill("wait_after_query_cache_invalidate"););
|
||||
|
||||
DBUG_VOID_RETURN;
|
||||
}
|
||||
|
||||
|
@ -904,20 +904,6 @@ bool mysql_insert(THD *thd,TABLE_LIST *table_list,
|
||||
}
|
||||
DBUG_ASSERT(transactional_table || !changed ||
|
||||
thd->transaction.stmt.modified_non_trans_table);
|
||||
|
||||
if (thd->lock)
|
||||
{
|
||||
/*
|
||||
Invalidate the table in the query cache if something changed
|
||||
after unlocking when changes become fisible.
|
||||
TODO: this is workaround. right way will be move invalidating in
|
||||
the unlock procedure.
|
||||
*/
|
||||
if (lock_type == TL_WRITE_CONCURRENT_INSERT && changed)
|
||||
{
|
||||
query_cache_invalidate3(thd, table_list, 1);
|
||||
}
|
||||
}
|
||||
}
|
||||
thd_proc_info(thd, "end");
|
||||
/*
|
||||
|
@ -2152,6 +2152,15 @@ my_bool ha_myisam::register_query_cache_table(THD *thd, char *table_name,
|
||||
}
|
||||
}
|
||||
|
||||
/*
|
||||
This query execution might have started after the query cache was flushed
|
||||
by a concurrent INSERT. In this case, don't cache this statement as the
|
||||
data file length difference might not be visible yet if the tables haven't
|
||||
been unlocked by the concurrent insert thread.
|
||||
*/
|
||||
if (file->state->uncacheable)
|
||||
DBUG_RETURN(FALSE);
|
||||
|
||||
/* It is ok to try to cache current statement. */
|
||||
DBUG_RETURN(TRUE);
|
||||
}
|
||||
|
@ -293,6 +293,8 @@ void mi_get_status(void* param, int concurrent_insert)
|
||||
info->save_state=info->s->state.state;
|
||||
info->state= &info->save_state;
|
||||
info->append_insert_at_end= concurrent_insert;
|
||||
if (concurrent_insert)
|
||||
info->s->state.state.uncacheable= TRUE;
|
||||
DBUG_VOID_RETURN;
|
||||
}
|
||||
|
||||
|
@ -38,6 +38,7 @@ typedef struct st_mi_status_info
|
||||
my_off_t key_file_length;
|
||||
my_off_t data_file_length;
|
||||
ha_checksum checksum;
|
||||
my_bool uncacheable; /* Active concurrent insert */
|
||||
} MI_STATUS_INFO;
|
||||
|
||||
typedef struct st_mi_state_info
|
||||
|
Loading…
x
Reference in New Issue
Block a user