1
0
mirror of https://github.com/MariaDB/server.git synced 2025-07-13 02:22:51 +03:00
Files
mariadb/storage/example/ha_example.h
Sven Sandberg f3985c649d BUG#39934: Slave stops for engine that only support row-based logging
General overview:
The logic for switching to row format when binlog_format=MIXED had
numerous flaws. The underlying problem was the lack of a consistent
architecture.
General purpose of this changeset:
This changeset introduces an architecture for switching to row format
when binlog_format=MIXED. It enforces the architecture where it has
to. It leaves some bugs to be fixed later. It adds extensive tests to
verify that unsafe statements work as expected and that appropriate
errors are produced by problems with the selection of binlog format.
It was not practical to split this into smaller pieces of work.

Problem 1:
To determine the logging mode, the code has to take several parameters
into account (namely: (1) the value of binlog_format; (2) the
capabilities of the engines; (3) the type of the current statement:
normal, unsafe, or row injection). These parameters may conflict in
several ways, namely:
 - binlog_format=STATEMENT for a row injection
 - binlog_format=STATEMENT for an unsafe statement
 - binlog_format=STATEMENT for an engine only supporting row logging
 - binlog_format=ROW for an engine only supporting statement logging
 - statement is unsafe and engine does not support row logging
 - row injection in a table that does not support statement logging
 - statement modifies one table that does not support row logging and
   one that does not support statement logging
Several of these conflicts were not detected, or were detected with
an inappropriate error message. The problem of BUG#39934 was that no
appropriate error message was written for the case when an engine
only supporting row logging executed a row injection with
binlog_format=ROW. However, all above cases must be handled.
Fix 1:
Introduce new error codes (sql/share/errmsg.txt). Ensure that all
conditions are detected and handled in decide_logging_format()

Problem 2:
The binlog format shall be determined once per statement, in
decide_logging_format(). It shall not be changed before or after that.
Before decide_logging_format() is called, all information necessary to
determine the logging format must be available. This principle ensures
that all unsafe statements are handled in a consistent way.
However, this principle is not followed:
thd->set_current_stmt_binlog_row_based_if_mixed() is called in several
places, including from code executing UPDATE..LIMIT,
INSERT..SELECT..LIMIT, DELETE..LIMIT, INSERT DELAYED, and
SET @@binlog_format. After Problem 1 was fixed, that caused
inconsistencies where these unsafe statements would not print the
appropriate warnings or errors for some of the conflicts.
Fix 2:
Remove calls to THD::set_current_stmt_binlog_row_based_if_mixed() from
code executed after decide_logging_format(). Compensate by calling the
set_current_stmt_unsafe() at parse time. This way, all unsafe statements
are detected by decide_logging_format().

Problem 3:
INSERT DELAYED is not unsafe: it is logged in statement format even if
binlog_format=MIXED, and no warning is printed even if
binlog_format=STATEMENT. This is BUG#45825.
Fix 3:
Made INSERT DELAYED set itself to unsafe at parse time. This allows
decide_logging_format() to detect that a warning should be printed or
the binlog_format changed.

Problem 4:
LIMIT clause were not marked as unsafe when executed inside stored
functions/triggers/views/prepared statements. This is
BUG#45785.
Fix 4:
Make statements containing the LIMIT clause marked as unsafe at
parse time, instead of at execution time. This allows propagating
unsafe-ness to the view.
2009-07-14 21:31:19 +02:00

252 lines
8.6 KiB
C++

/* Copyright (C) 2003 MySQL AB
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; version 2 of the License.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */
/** @file ha_example.h
@brief
The ha_example engine is a stubbed storage engine for example purposes only;
it does nothing at this point. Its purpose is to provide a source
code illustration of how to begin writing new storage engines; see also
/storage/example/ha_example.cc.
@note
Please read ha_example.cc before reading this file.
Reminder: The example storage engine implements all methods that are *required*
to be implemented. For a full list of all methods that you can implement, see
handler.h.
@see
/sql/handler.h and /storage/example/ha_example.cc
*/
#ifdef USE_PRAGMA_INTERFACE
#pragma interface /* gcc class implementation */
#endif
/** @brief
EXAMPLE_SHARE is a structure that will be shared among all open handlers.
This example implements the minimum of what you will probably need.
*/
typedef struct st_example_share {
char *table_name;
uint table_name_length,use_count;
pthread_mutex_t mutex;
THR_LOCK lock;
} EXAMPLE_SHARE;
/** @brief
Class definition for the storage engine
*/
class ha_example: public handler
{
THR_LOCK_DATA lock; ///< MySQL lock
EXAMPLE_SHARE *share; ///< Shared lock info
public:
ha_example(handlerton *hton, TABLE_SHARE *table_arg);
~ha_example()
{
}
/** @brief
The name that will be used for display purposes.
*/
const char *table_type() const { return "EXAMPLE"; }
/** @brief
The name of the index type that will be used for display.
Don't implement this method unless you really have indexes.
*/
const char *index_type(uint inx) { return "HASH"; }
/** @brief
The file extensions.
*/
const char **bas_ext() const;
/** @brief
This is a list of flags that indicate what functionality the storage engine
implements. The current table flags are documented in handler.h
*/
ulonglong table_flags() const
{
/*
We are saying that this engine is just statement capable to have
an engine that can only handle statement-based logging. This is
used in testing.
*/
return HA_BINLOG_STMT_CAPABLE;
}
/** @brief
This is a bitmap of flags that indicates how the storage engine
implements indexes. The current index flags are documented in
handler.h. If you do not implement indexes, just return zero here.
@details
part is the key part to check. First key part is 0.
If all_parts is set, MySQL wants to know the flags for the combined
index, up to and including 'part'.
*/
ulong index_flags(uint inx, uint part, bool all_parts) const
{
return 0;
}
/** @brief
unireg.cc will call max_supported_record_length(), max_supported_keys(),
max_supported_key_parts(), uint max_supported_key_length()
to make sure that the storage engine can handle the data it is about to
send. Return *real* limits of your storage engine here; MySQL will do
min(your_limits, MySQL_limits) automatically.
*/
uint max_supported_record_length() const { return HA_MAX_REC_LENGTH; }
/** @brief
unireg.cc will call this to make sure that the storage engine can handle
the data it is about to send. Return *real* limits of your storage engine
here; MySQL will do min(your_limits, MySQL_limits) automatically.
@details
There is no need to implement ..._key_... methods if your engine doesn't
support indexes.
*/
uint max_supported_keys() const { return 0; }
/** @brief
unireg.cc will call this to make sure that the storage engine can handle
the data it is about to send. Return *real* limits of your storage engine
here; MySQL will do min(your_limits, MySQL_limits) automatically.
@details
There is no need to implement ..._key_... methods if your engine doesn't
support indexes.
*/
uint max_supported_key_parts() const { return 0; }
/** @brief
unireg.cc will call this to make sure that the storage engine can handle
the data it is about to send. Return *real* limits of your storage engine
here; MySQL will do min(your_limits, MySQL_limits) automatically.
@details
There is no need to implement ..._key_... methods if your engine doesn't
support indexes.
*/
uint max_supported_key_length() const { return 0; }
/** @brief
Called in test_quick_select to determine if indexes should be used.
*/
virtual double scan_time() { return (double) (stats.records+stats.deleted) / 20.0+10; }
/** @brief
This method will never be called if you do not implement indexes.
*/
virtual double read_time(ha_rows rows) { return (double) rows / 20.0+1; }
/*
Everything below are methods that we implement in ha_example.cc.
Most of these methods are not obligatory, skip them and
MySQL will treat them as not implemented
*/
/** @brief
We implement this in ha_example.cc; it's a required method.
*/
int open(const char *name, int mode, uint test_if_locked); // required
/** @brief
We implement this in ha_example.cc; it's a required method.
*/
int close(void); // required
/** @brief
We implement this in ha_example.cc. It's not an obligatory method;
skip it and and MySQL will treat it as not implemented.
*/
int write_row(uchar *buf);
/** @brief
We implement this in ha_example.cc. It's not an obligatory method;
skip it and and MySQL will treat it as not implemented.
*/
int update_row(const uchar *old_data, uchar *new_data);
/** @brief
We implement this in ha_example.cc. It's not an obligatory method;
skip it and and MySQL will treat it as not implemented.
*/
int delete_row(const uchar *buf);
/** @brief
We implement this in ha_example.cc. It's not an obligatory method;
skip it and and MySQL will treat it as not implemented.
*/
int index_read_map(uchar *buf, const uchar *key,
key_part_map keypart_map, enum ha_rkey_function find_flag);
/** @brief
We implement this in ha_example.cc. It's not an obligatory method;
skip it and and MySQL will treat it as not implemented.
*/
int index_next(uchar *buf);
/** @brief
We implement this in ha_example.cc. It's not an obligatory method;
skip it and and MySQL will treat it as not implemented.
*/
int index_prev(uchar *buf);
/** @brief
We implement this in ha_example.cc. It's not an obligatory method;
skip it and and MySQL will treat it as not implemented.
*/
int index_first(uchar *buf);
/** @brief
We implement this in ha_example.cc. It's not an obligatory method;
skip it and and MySQL will treat it as not implemented.
*/
int index_last(uchar *buf);
/** @brief
Unlike index_init(), rnd_init() can be called two consecutive times
without rnd_end() in between (it only makes sense if scan=1). In this
case, the second call should prepare for the new table scan (e.g if
rnd_init() allocates the cursor, the second call should position the
cursor to the start of the table; no need to deallocate and allocate
it again. This is a required method.
*/
int rnd_init(bool scan); //required
int rnd_end();
int rnd_next(uchar *buf); ///< required
int rnd_pos(uchar *buf, uchar *pos); ///< required
void position(const uchar *record); ///< required
int info(uint); ///< required
int extra(enum ha_extra_function operation);
int external_lock(THD *thd, int lock_type); ///< required
int delete_all_rows(void);
ha_rows records_in_range(uint inx, key_range *min_key,
key_range *max_key);
int delete_table(const char *from);
int rename_table(const char * from, const char * to);
int create(const char *name, TABLE *form,
HA_CREATE_INFO *create_info); ///< required
THR_LOCK_DATA **store_lock(THD *thd, THR_LOCK_DATA **to,
enum thr_lock_type lock_type); ///< required
};