GIF89a; Mini Shell

Mini Shell

Direktori : /usr/share/mysql-test/extra/rpl_tests/
Upload File :
Current File : //usr/share/mysql-test/extra/rpl_tests/rpl_insert_id.test

###########################################################
# See if queries that use both auto_increment and LAST_INSERT_ID()
# are replicated well
############################################################
# REQUIREMENT
# Auto increment should work for a table with an auto_increment
# column and index but without primary key.
##############################################################

--echo #
--echo # Setup
--echo #

--echo #
--echo # See if queries that use both auto_increment and LAST_INSERT_ID()
--echo # are replicated well
--echo #
--echo # We also check how the foreign_key_check variable is replicated
--echo #

-- source include/master-slave.inc
#should work for both SBR and RBR

--disable_query_log
call mtr.add_suppression("Unsafe statement written to the binary log using statement format since BINLOG_FORMAT = STATEMENT");
--connection slave
call mtr.add_suppression("Unsafe statement written to the binary log using statement format since BINLOG_FORMAT = STATEMENT");
--enable_query_log


# If concurrent inserts are on, it is not guaranteed that the rows
# inserted by INSERT are immediately accessible by SELECT in another
# thread. This would cause problems near the line 'connection master1'
# below. So we turn off concurrent inserts.
connection master;
SET @old_concurrent_insert= @@global.concurrent_insert;
SET @@global.concurrent_insert= 0;

connection master;
eval create table t1(a int auto_increment, key(a)) engine=$engine_type;
eval create table t2(b int auto_increment, c int, key(b)) engine=$engine_type;
insert into t1 values (1),(2),(3);
insert into t1 values (null);
insert into t2 values (null,last_insert_id());
--source include/sync_slave_sql_with_master.inc
select * from t1 ORDER BY a;
select * from t2 ORDER BY b;
connection master;
#check if multi-line inserts,
#which set last_insert_id to the first id inserted,
#are replicated the same way
drop table t1;
drop table t2;
--disable_warnings
eval create table t1(a int auto_increment, key(a)) engine=$engine_type;
eval create table t2(b int auto_increment, c int, key(b), foreign key(b) references t1(a)) engine=$engine_type;
--enable_warnings
SET FOREIGN_KEY_CHECKS=0;
insert into t1 values (10);
insert into t1 values (null),(null),(null);
insert into t2 values (5,0);
insert into t2 values (null,last_insert_id());
SET FOREIGN_KEY_CHECKS=1;
--source include/sync_slave_sql_with_master.inc
select * from t1;
select * from t2;
connection master;

--echo #
--echo # check if INSERT SELECT in auto_increment is well replicated (bug #490)
--echo #

drop table t2;
drop table t1;
eval create table t1(a int auto_increment, key(a)) engine=$engine_type;
eval create table t2(b int auto_increment, c int, key(b)) engine=$engine_type;
insert into t1 values (10);
insert into t1 values (null),(null),(null);
insert into t2 values (5,0);
--disable_warnings ONCE
insert into t2 (c) select * from t1 ORDER BY a;
select * from t2 ORDER BY b;
--source include/sync_slave_sql_with_master.inc
select * from t1 ORDER BY a;
select * from t2 ORDER BY b;
connection master;
drop table t1;
drop table t2;
--source include/sync_slave_sql_with_master.inc

--echo #
--echo # Bug#8412: Error codes reported in binary log for CHARACTER SET,
--echo #           FOREIGN_KEY_CHECKS
--echo #

connection master;
SET TIMESTAMP=1000000000;
eval CREATE TABLE t1 ( a INT UNIQUE ) engine=$engine_type;
SET FOREIGN_KEY_CHECKS=0;
# Duplicate Key Errors
--error 1022, ER_DUP_ENTRY
INSERT INTO t1 VALUES (1),(1);
--source include/sync_slave_sql_with_master.inc
connection master;
drop table t1;
--source include/sync_slave_sql_with_master.inc

--echo #
--echo # Bug#14553: NULL in WHERE resets LAST_INSERT_ID
--echo #

connection master;
set @@session.sql_auto_is_null=1;
eval create table t1(a int auto_increment, key(a)) engine=$engine_type;
eval create table t2(a int) engine=$engine_type;
insert into t1 (a) values (null);
--disable_warnings
insert into t2 (a) select a from t1 where a is null;
insert into t2 (a) select a from t1 where a is null;
--enable_warnings
select * from t2;
--source include/sync_slave_sql_with_master.inc
connection slave;
select * from t2;
connection master;
drop table t1;
drop table t2;

--echo #
--echo # End of 4.1 tests
--echo #

--echo #
--echo # BUG#15728: LAST_INSERT_ID function inside a stored function returns 0
--echo #
--echo # The solution is not to reset last_insert_id on enter to sub-statement.
--echo #

connection master;
--disable_warnings
drop function if exists bug15728;
drop function if exists bug15728_insert;
drop table if exists t1, t2;
--enable_warnings

eval create table t1 (
  id int not null auto_increment,
  last_id int,
  primary key (id)
) engine=$engine_type;
create function bug15728() returns int(11)
  return last_insert_id();

insert into t1 (last_id) values (0);
insert into t1 (last_id) values (last_insert_id());
insert into t1 (last_id) values (bug15728());

# Check that nested call replicates too.
eval create table t2 (
  id int not null auto_increment,
  last_id int,
  primary key (id)
) engine=$engine_type;
delimiter |;
create function bug15728_insert() returns int(11) modifies sql data
begin
  insert into t2 (last_id) values (bug15728());
  return bug15728();
end|
create trigger t1_bi before insert on t1 for each row
begin
  declare res int;
  select bug15728_insert() into res;
  set NEW.last_id = res;
end|
delimiter ;|

--disable_warnings ONCE
insert into t1 (last_id) values (0);

drop trigger t1_bi;

# Check that nested call doesn't affect outer context.
select last_insert_id();
--disable_warnings ONCE
select bug15728_insert();
select last_insert_id();
insert into t1 (last_id) values (bug15728());
# This should be exactly one greater than in the previous call.
select last_insert_id();

# BUG#20339 - stored procedure using LAST_INSERT_ID() does not
# replicate statement-based
--disable_warnings ONCE
drop procedure if exists foo;
delimiter |;
create procedure foo()
begin
  declare res int;
  insert into t2 (last_id) values (bug15728());
  insert into t1 (last_id) values (bug15728());
end|
delimiter ;|
call foo();

select * from t1;
select * from t2;
--source include/sync_slave_sql_with_master.inc
select * from t1;
select * from t2;
connection master;

drop function bug15728;
drop function bug15728_insert;
drop table t1,t2;
drop procedure foo;

# test of BUG#20188 REPLACE or ON DUPLICATE KEY UPDATE in
# auto_increment breaks binlog

eval create table t1 (n int primary key auto_increment not null,
                      b int, unique(b)) engine=$engine_type;

# First, test that we do not call restore_auto_increment() too early
# in write_record():
set sql_log_bin=0;
insert into t1 values(null,100);
replace into t1 values(null,50),(null,100),(null,150);
select * from t1 order by n;
truncate table t1;
set sql_log_bin=1;

insert into t1 values(null,100);
select * from t1 order by n;
--source include/sync_slave_sql_with_master.inc
# make slave's table autoinc counter bigger
insert into t1 values(null,200),(null,300);
delete from t1 where b <> 100;
# check that slave's table content is identical to master
select * from t1 order by n;
# only the auto_inc counter differs.

connection master;
replace into t1 values(null,100),(null,350);
select * from t1 order by n;
--source include/sync_slave_sql_with_master.inc
select * from t1 order by n;

# Same test as for REPLACE, but for ON DUPLICATE KEY UPDATE

# We first check that if we update a row using a value larger than the
# table's counter, the counter for next row is bigger than the
# after-value of the updated row.
connection master;
--disable_warnings ONCE
insert into t1 values (NULL,400),(3,500),(NULL,600) on duplicate key UPDATE n=1000;
select * from t1 order by n;
--source include/sync_slave_sql_with_master.inc
select * from t1 order by n;

# and now test for the bug:
connection master;
drop table t1;
eval create table t1 (n int primary key auto_increment not null,
                      b int, unique(b)) engine=$engine_type;
insert into t1 values(null,100);
select * from t1 order by n;
--source include/sync_slave_sql_with_master.inc
insert into t1 values(null,200),(null,300);
delete from t1 where b <> 100;
select * from t1 order by n;

connection master;
--disable_warnings ONCE
insert into t1 values(null,100),(null,350) on duplicate key update n=2;
select * from t1 order by n;
--source include/sync_slave_sql_with_master.inc
select * from t1 order by n;

connection master;
drop table t1;
--source include/sync_slave_sql_with_master.inc

#
# BUG#24432 "INSERT... ON DUPLICATE KEY UPDATE skips auto_increment values"
#

connection master;
# testcase with INSERT VALUES
eval CREATE TABLE t1 (a INT NOT NULL PRIMARY KEY AUTO_INCREMENT, b INT,
                      UNIQUE(b)) ENGINE=$engine_type;
--disable_warnings ONCE
INSERT INTO t1(b) VALUES(1),(1),(2) ON DUPLICATE KEY UPDATE t1.b=10;
SELECT * FROM t1 ORDER BY a;
--source include/sync_slave_sql_with_master.inc
SELECT * FROM t1 ORDER BY a;
connection master;
drop table t1;

# tescase with INSERT SELECT
eval CREATE TABLE t1 (
  id bigint(20) unsigned NOT NULL auto_increment,
  field_1 int(10) unsigned NOT NULL,
  field_2 varchar(255) NOT NULL,
  field_3 varchar(255) NOT NULL,
  PRIMARY KEY (id),
  UNIQUE KEY field_1 (field_1, field_2)
) ENGINE=$engine_type;
eval CREATE TABLE t2 (
  field_a int(10) unsigned NOT NULL,
  field_b varchar(255) NOT NULL,
  field_c varchar(255) NOT NULL
) ENGINE=$engine_type;
INSERT INTO t2 (field_a, field_b, field_c) VALUES (1, 'a', '1a');
INSERT INTO t2 (field_a, field_b, field_c) VALUES (2, 'b', '2b');
INSERT INTO t2 (field_a, field_b, field_c) VALUES (3, 'c', '3c');
INSERT INTO t2 (field_a, field_b, field_c) VALUES (4, 'd', '4d');
INSERT INTO t2 (field_a, field_b, field_c) VALUES (5, 'e', '5e');
# Updating table t1 based on values from table t2
--disable_warnings
INSERT INTO t1 (field_1, field_2, field_3)
SELECT t2.field_a, t2.field_b, t2.field_c
FROM t2
ON DUPLICATE KEY UPDATE
t1.field_3 = t2.field_c;
--enable_warnings
# Inserting new record into t2
INSERT INTO t2 (field_a, field_b, field_c) VALUES (6, 'f', '6f');
# Updating t1 again
--disable_warnings
INSERT INTO t1 (field_1, field_2, field_3)
SELECT t2.field_a, t2.field_b, t2.field_c
FROM t2
ON DUPLICATE KEY UPDATE
t1.field_3 = t2.field_c;
--enable_warnings
SELECT * FROM t1 ORDER BY id;
--source include/sync_slave_sql_with_master.inc
SELECT * FROM t1 ORDER BY id;
connection master;
drop table t1, t2;

#
# BUG#20339: stored procedure using LAST_INSERT_ID() does not
# replicate statement-based.
#
# There is another version of the test for bug#20339 above that is
# actually originates in 5.1, and this is the version that is merged
# from 5.0.
#
connection master;

--disable_warnings
DROP PROCEDURE IF EXISTS p1;
DROP TABLE IF EXISTS t1, t2;
--enable_warnings

# Reset result of LAST_INSERT_ID().
SELECT LAST_INSERT_ID(0);

eval CREATE TABLE t1 (
  id INT NOT NULL DEFAULT 0,
  last_id INT,
  PRIMARY KEY (id)
) ENGINE=$engine_type;

eval CREATE TABLE t2 (
  id INT NOT NULL AUTO_INCREMENT,
  last_id INT,
  PRIMARY KEY (id)
) ENGINE=$engine_type;

delimiter |;
CREATE PROCEDURE p1()
BEGIN
  INSERT INTO t2 (last_id) VALUES (LAST_INSERT_ID());
  INSERT INTO t1 (last_id) VALUES (LAST_INSERT_ID());
END|
delimiter ;|

CALL p1();
SELECT * FROM t1 ORDER BY id;
SELECT * FROM t2 ORDER BY id;

--source include/sync_slave_sql_with_master.inc
SELECT * FROM t1 ORDER BY id;
SELECT * FROM t2 ORDER BY id;

connection master;

DROP PROCEDURE p1;
DROP TABLE t1, t2;


#
# BUG#21726: Incorrect result with multiple invocations of
# LAST_INSERT_ID
#
--disable_warnings
DROP PROCEDURE IF EXISTS p1;
DROP FUNCTION IF EXISTS f1;
DROP FUNCTION IF EXISTS f2;
DROP FUNCTION IF EXISTS f3;
DROP TABLE IF EXISTS t1, t2;
--enable_warnings

eval CREATE TABLE t1 (
    i INT NOT NULL AUTO_INCREMENT PRIMARY KEY,
    j INT DEFAULT 0
) ENGINE=$engine_type;
eval CREATE TABLE t2 (i INT) ENGINE=$engine_type;

delimiter |;
CREATE PROCEDURE p1()
BEGIN
  INSERT INTO t1 (i) VALUES (NULL);
  INSERT INTO t2 (i) VALUES (LAST_INSERT_ID());
  INSERT INTO t1 (i) VALUES (NULL), (NULL);
  INSERT INTO t2 (i) VALUES (LAST_INSERT_ID());
END |

CREATE FUNCTION f1() RETURNS INT MODIFIES SQL DATA
BEGIN
  INSERT INTO t1 (i) VALUES (NULL);
  INSERT INTO t2 (i) VALUES (LAST_INSERT_ID());
  INSERT INTO t1 (i) VALUES (NULL), (NULL);
  INSERT INTO t2 (i) VALUES (LAST_INSERT_ID());
  RETURN 0;
END |

CREATE FUNCTION f2() RETURNS INT NOT DETERMINISTIC
  RETURN LAST_INSERT_ID() |

CREATE FUNCTION f3() RETURNS INT MODIFIES SQL DATA
BEGIN
  INSERT INTO t2 (i) VALUES (LAST_INSERT_ID());
  RETURN 0;
END |
delimiter ;|

INSERT INTO t1 VALUES (NULL, -1);
CALL p1();
--disable_warnings ONCE
SELECT f1();
INSERT INTO t1 VALUES (NULL, f2()), (NULL, LAST_INSERT_ID()),
                      (NULL, LAST_INSERT_ID()), (NULL, f2()), (NULL, f2());
INSERT INTO t1 VALUES (NULL, f2());
# Test replication of substitution "IS NULL" -> "= LAST_INSERT_ID".
INSERT INTO t1 VALUES (NULL, 0), (NULL, LAST_INSERT_ID());
UPDATE t1 SET j= -1 WHERE i IS NULL;

# Test statement-based replication of function calls.
INSERT INTO t1 (i) VALUES (NULL);

# Here, we rely on having set @@concurrent_insert= 0 (see comment at
# the top of this file).
connection master1;
INSERT INTO t1 (i) VALUES (NULL);

connection master;
SELECT f3();

SELECT * FROM t1 ORDER BY i;
SELECT * FROM t2 ORDER BY i;

--source include/sync_slave_sql_with_master.inc
SELECT * FROM t1;
SELECT * FROM t2;

connection master;
DROP PROCEDURE p1;
DROP FUNCTION f1;
DROP FUNCTION f2;
DROP FUNCTION f3;
DROP TABLE t1, t2;


--source include/sync_slave_sql_with_master.inc

--echo #
--echo # End of 5.0 tests
--echo #

# Tests in this file are tightly bound together.  Recreate t2.
connection master;
eval create table t2 (
  id int not null auto_increment,
  last_id int,
  primary key (id)
) engine=$engine_type;


# Test for BUG#20341 "stored function inserting into one
# auto_increment puts bad data in slave"

connection master;
truncate table t2;
# no auto_increment
eval create table t1 (id tinyint primary key) engine=$engine_type;

delimiter |;
create function insid() returns int
begin
  insert into t2 (last_id) values (0);
  return 0;
end|
delimiter ;|
set sql_log_bin=0;
insert into t2 (id) values(1),(2),(3);
delete from t2;
set sql_log_bin=1;
#inside SELECT, then inside INSERT
--disable_warnings ONCE
select insid();
set sql_log_bin=0;
insert into t2 (id) values(5),(6),(7);
delete from t2 where id>=5;
set sql_log_bin=1;
--disable_warnings ONCE
insert into t1 select insid();
select * from t1 order by id;
select * from t2 order by id;

--source include/sync_slave_sql_with_master.inc
select * from t1 order by id;
select * from t2 order by id;

connection master;
drop table t1;
drop function insid;

truncate table t2;
eval create table t1 (n int primary key auto_increment not null,
                      b int, unique(b)) engine=$engine_type;
delimiter |;
create procedure foo()
begin
  insert into t1 values(null,10);
  insert ignore into t1 values(null,10);
  insert ignore into t1 values(null,10);
  insert into t2 values(null,3);
end|
delimiter ;|
--disable_warnings ONCE
call foo();
select * from t1 order by n;
select * from t2 order by id;

--source include/sync_slave_sql_with_master.inc
select * from t1 order by n;
select * from t2 order by id;

connection master;
drop table t1, t2;
drop procedure foo;
SET @@global.concurrent_insert= @old_concurrent_insert;
set @@session.sql_auto_is_null=default;

--source include/rpl_end.inc

./BlackJoker Mini Shell 1.0